UNIVERSIDADE FEDERAL DO RIO GRANDE DO NORTE CENTRO DE CIÊNCIAS EXATAS E DA TERRA DEPARTAMENTO DE INFORMÁTICA E MATEMÁTICA APLICADA PROGRAMA DE PÓS-GRADUAÇÃO EM SISTEMAS E COMPUTAÇÃO MESTRADO ACADÊMICO EM SISTEMAS E COMPUTAÇÃO

Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais: Estudo e Desenvolvimento do Jogo Karuchā Ships Invaders

Juvane Nunes Marciano

Natal-RN Fevereiro de 2014

Juvane Nunes Marciano

Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais: Estudo e Desenvolvimento do Jogo Karuchā Ships Invaders

Dissertação de Mestrado apresentado ao Programa de Pós-Graduação em Sistemas e Computação do Departamento de Informática e Matemática Aplicada da Universidade Federal do Rio Grande do Norte como requisito final para a obtenção do grau de Mestre em Sistemas e Computação.

Linha de pesquisa: Engenharia de Software

Orientador Prof. Dr. Leonardo Cunha de Miranda

PPGSC – PROGRAMA DE PÓS-GRADUAÇÃO EM SISTEMAS E COMPUTAÇÃO DIMAP – DEPARTAMENTO DE INFORMÁTICA E MATEMÁTICA APLICADA CCET – CENTRO DE CIÊNCIAS EXATAS E DA TERRA UFRN – UNIVERSIDADE FEDERAL DO RIO GRANDE DO NORTE

Natal-RN Fevereiro de 2014

UFRN / Biblioteca Central Zila Mamede Catalogação da Publicação na Fonte

Marciano, Juvane Nunes. Aprendizagem da língua japonesa apoiada por ferramentas computacionais: estudo e desenvolvimento do jogo Karuchā Ships Invaders. / Juvane Nunes Marciano. – Natal, RN, 2014. 140 f.: il.

Orientador: Prof. Dr. Leonardo Cunha de Miranda.

Dissertação (Mestrado) – Universidade Federal do Rio Grande do Norte. Centro de Ciências Exatas e da Terra. Programa de Pós-Graduação em Sistemas e Computação.

1. Informática na educação – Dissertação. 2. Interação humano- computador - Dissertação. 3. Jogo educacional - Dissertação. 4. Kanas – Dissertação. 5. Hiragana – Dissertação. 6. Roma-ji – Dissertação. I. Miranda, Leonardo Cunha de. II. Universidade Federal do Rio Grande do Norte. III. Título.

RN/UF/BCZM CDU 004:37

Dissertação de Mestrado sob o título Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais: Estudo e Desenvolvimento do Jogo Karuchā Ships Invaders, apresentada por Juvane Nunes Marciano e aceita pelo Programa de Pós-Graduação em Sistemas e Computação do Departamento de Informática e Matemática Aplicada da Universidade Federal do Rio Grande do Norte, sendo aprovada por todos os membros da banca examinadora abaixo especificada:

______Prof. Dr. Leonardo Cunha de Miranda Presidente Departamento de Informática e Matemática Aplicada (DIMAp) Universidade Federal do Rio Grande do Norte (UFRN)

______Prof. Dr. Fábio Ferrentini Sampaio Examinador Externo Instituto Tércio Pacitti de Aplicações e Pesquisas Computacionais (iNCE) Universidade Federal do Rio de Janeiro (UFRJ)

______Prof. Dr. André Maurício Cunha Campos Examinador Interno Departamento de Informática e Matemática Aplicada (DIMAp) Universidade Federal do Rio Grande do Norte (UFRN)

______Prof. Dra. Márcia Jacyntha Nunes Rodrigues Lucena Examinador Interno Departamento de Informática e Matemática Aplicada (DIMAp) Universidade Federal do Rio Grande do Norte (UFRN)

Natal-RN, 24, Fevereiro de 2014

À minha mãe

Agradecimentos

Agradeço primeiramente a Deus, por tudo, especialmente pelo Sábado, o dia para revigorar as energias a cada semana e poder continuar o trabalho com fé, força e determinação.

Agradeço à minha família (inteira, em todo canto do mundo), especialmente minha mãe, Juciane, por todo o apoio durante mais essa jornada de estudos, e ao meu irmão e minha cunhada, que são meu lar na cidade de Natal. Também agradeço ao Nando, pelo mega apoio agora na reta final da jornada.

Agradeço à minha cidade, Currais Novos, pelos fins de semana bem gastos – estudando ou aproveitando –, e aos amigos e irmãos que me apoiaram durante a jornada (são muitos para citar no pouco espaço, então, sintam-se todos inclusos).

Agradeço ao meu orientador, o professor Leonardo, pela incontestável dedicação a todos os seus alunos, incluindo a mim, pela energia e o otimismo no cumprimento das atividades, pela compreensão nos momentos em que tudo ficava mais difícil, pelo esforço hercúleo aplicado para que tudo no trabalho “faça sentido” e dentro dos prazos, pelo acompanhamento assíduo que sempre ajudou a despertar o esforço e as ideias para seguir com a pesquisa. E, por fim, por ter se preocupado em me conhecer, descobrir algo que para mim era um hobby e transformar em uma pesquisa científica que eu nem sabia que era possível desenvolver.

Agradeço à Universidade Federal do Rio Grande do Norte – UFRN – e ao Departamento de Informática e Matemática Aplicada – DIMAp – pela oportunidade de cursar a pós graduação em seu programa.

Agradeço ao CNPq pela bolsa de estudos a mim concedida.

Agradeço ao grupo Physical Artifacts of Interaction Research Group – PAIRG – como um todo, pois cada ideia, crítica e sugestão, de cada membro, contribuiu para o enriquecimento desta pesquisa. Em especial à professora Erica, que sempre contribuiu muito significativamente em todas as etapas deste trabalho.

Agradeço à equipe de desenvolvimento do jogo Karuchā Ships Invaders, por ter tornado possível a construção dessa ferramenta educacional que se mostrou com grande potencial de crescimento graças ao esforço de todos os envolvidos. Também agradeço aos doze participantes do experimento com usuário realizado para avaliação do Karuchā.

Agradeço aos membros da banca avaliadora, professor Fábio, professor André e professora Márcia, por terem aceitado o convite para contribuir com este trabalho. Também pelas observações importantíssimas feitas pelos professores André e Márcia durante minha banca de qualificação.

Agradeço, por fim, a todos os demais que de alguma forma, qualquer forma, contribuíram para a realização desta pesquisa.

“For I have always been a seeker, a dreamer, and a ponderer on seeking and dreaming...” ― H.P. Lovecraft

Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais: Estudo e Desenvolvimento do Jogo Karuchā Ships Invaders

Autora: Juvane Nunes Marciano Orientador: Prof. Dr. Leonardo Cunha de Miranda

RESUMO

O uso de ferramentas tecnológicas para apoiar o processo de ensino-aprendizagem de diversas áreas do conhecimento é notório e vem crescendo a cada dia, inclusive, no contexto de aprendizado de idiomas. O estudo de Aprendizado de Idioma com Auxílio de Computador – Computer Assisted Language Learning (CALL) – realizado nesta pesquisa teve o intuito de avaliar as ferramentas existentes nesse contexto, focadas especificamente no idioma japonês e, a partir desse estudo, realizar o desenvolvimento de uma nova ferramenta que auxilie o ensino da língua japonesa. Como resultados, apresentamos um abrangente levantamento sobre o tema nas mais diversas tecnologias, bem como, descrevemos todos os passos do desenvolvimento do jogo Karuchā Ships Invaders, que se propõe a ensinar conceitos básicos do idioma, mesclando-se ao entretenimento, ainda com foco no público de estudantes brasileiros de japonês. Serão apresentadas as fases de concepção do jogo e sua evolução ao longo da pesquisa, bem como a avaliação de sua interface. Ainda, apresentamos um método para avaliar aspectos motivacionais de ferramentas computacionais com foco educacional e resultados de um experimento realizado com usuários em perspectiva.

Palavras-chave: Informática na Educação, Interação Humano-Computador, Jogo Educacional, Kanas, Hiragana, Roma-ji.

Japanese Language Learning Supported by Computational Tools: Study and Development of Karuchā Ships Invaders Game

Author: Juvane Nunes Marciano Advisor: Prof. Dr. Leonardo Cunha de Miranda

ABSTRACT

The use of technology tools for teaching and learning has grown increasingly in our daily life. In this context, a branch that has had tremendous growth is the area of teaching and learning language through computational tools. The study of CALL (Computer Assisted Language Learning), accomplished in this research, aims to evaluate existing tools in this context, focused specifically on the Japanese language; and from this study, accomplish the development of a new computational tool that can assist teaching/learning of the Japanese language. As results, we present a wide survey on the subject in various technologies/devices, as well as the complete development process of a new tool, the Karuchā Ships Invaders game, that proposes to teach basic concepts of the language, blended with entertainment, and still, focusing on the Brazilian students of Japanese language audience. We will present all the concept phases of the game and its evolution through the research, as well as an interface evaluation. Still, we present proposal and validation of a method to evaluate motivational aspects of computational tools with educational focus, and results extracted from an experiment accomplished with prospective users.

Keywords: Informatics in Education; Human-Computer Interaction; Educational Game; Kanas; Hiragana; Roma-ji.

Lista de figuras

Fig. 1 - Origem pictográfica do kanji de “montanha”...... 22 Fig. 2 - Junção de radicais e ideias para formar um novo kanji...... 23 Fig. 3 - Uso de tenten, maru, e junção de caracteres...... 24 Fig. 4 - América do Sul e América Latina escritos em (a) hiragana (b) kanji (c) roma-ji (d) katakana...... 24 Fig. 5 - The semiotic framework (c.f. [40])...... 57 Fig. 6 - Reading skill quiz (Elementary Kanji App)...... 69 Fig. 7 - Japanese text with roma-ji and translation, audio teaching, and user speaking recognition (Japanese Conversation Lite App)...... 69 Fig. 8 - Kanji drawing recognition (Kanji Recognizer App)...... 69 Fig. 9 - Animation showing the stoke order for the hiragana “ga” (Simple Japanese Gojuon App)...... 70 Fig. 10 - Listening skill quiz (Learn Japanese Quick App)...... 70 Fig. 11 - Main menu interface of the game...... 86 Fig. 12 - Level and difficulty mode interface of the game...... 86 Fig. 13 - Scenario transitions of the game, i.e. from/to (a) morning, (b) evening, and (c) night. .. 87 Fig. 14 - Ships of the game for (a) hiragana character (b) cultural element...... 87 Fig. 15 - Japanese character that represents the player’s performance on a given level, i.e. (a) passed level with bonus (b) passed level (c) game over...... 88 Fig. 16 - Components of the game (a) hit points bar (b) time (c) difficulty mode (d) level of the game (e) ship of hiragana (f) ship of cultural aspect (g) guidance laser ray (h) city (i) laser cannon (j) feedback area of user’s typing...... 88 Fig. 17 - Different moments of the Karuchā Ships Invaders CALL game...... 95 Fig. 18 - Five steps of the evaluation method...... 100 Fig. 19 - Steps of the method...... 110 Fig. 20 - Diagrams that detail user characteristics...... 120 Fig. 21 - Maximum level the users achieved...... 121 Fig. 22 - Counter of use of the context-help in each level...... 122 Fig. 23 - A flow through levels of hard mode, from user “U3”...... 122 Fig. 24 - Severity table of the motivational questionnaire’s answers...... 123 Fig. 25 - Users playing Karuchā Ships Invaders...... 124

Fig. 26 - Users’ reactions to the game...... 124 Fig. 27 - Users answering motivational questionnaire and interview...... 125

Lista de tabelas

Table 1 - Foco dos Trabalhos da Literatura...... 35 Table 2 - Medias used by CALL Web Tools for the Japanese language, identified by language skills...... 49 Table 3 - State of the Practice Tools, synthesizing media behavior when accessed from the Android platform...... 50 Table 4 - Android MALL Apps Identified and Evaluated ...... 65 Table 5 - Semiotic Framework in the Analysis Domain of MALL Apps for Japanese...... 67 Table 6 - Identified Requirements According to Apps’ Classification ...... 71 Table 7 - Related Works and Games...... 78 Table 8 - Basic Elements of the Game ...... 81 Table 9 - Some Examples of Bosses of the Game...... 83 Table 10 - Achievements of the Game...... 84 Table 11 - Summary of the works of literature review...... 99 Table 12 - Basis heuristics for the development of the questions...... 101 Table 13 - Questions defined to assess multiple interface aspects of Karuchā Ships Invaders. .. 101 Table 14 - Answers for the heuristic based questions...... 102 Table 15 - Severity rating of problems found...... 104 Table 16 - Map set of heuristics developed for use with the method...... 111 Table 17 - Example questions guided by the heuristics for the example of the method...... 113 Table 18 - Examples of motivational strengths and weaknesses...... 114

Lista de abreviaturas e siglas

ARCS: Attention, Relevance, Confidence, Satisfaction CALL: Computer Assisted Language Learning FAQ: Frequently Asked Questions GPS: Global Positioning System HCI: Human-Computer Interaction HKR: Handwritten Kanji Recognition IHC: Interação Humano-Computador IMMS: Instructional Materials Motivational Survey IT: Information Technology JAMIOLAS: Japanese Mimicry and Onomatopoeia Learning Assistant JAPELAS: Japanese Polite Expressions Learning Assisting System JLPT: Japanese Language Proficiency Test JND: Japanese Names Dictionary JVC: Japanese Verb Conjugator JVD: Japanese Verb Deconjugator LAN: Local Area Network MALL: Mobile Assisted Language Learning MAWL: Mobile Assisted Word Learning MIO: Mímica e Onomatopeias NHK: Nippon Hōsō Kyōkai OS: Operating System /or/ Organizational Semiotics P&D: Pesquisa e Desenvolvimento PATS: Proverbial Animation of Ten Sounds PC: Personal Computer PDA: Personal Digital Assistant PEFS: Proverbial Exam by Fourfold Selection PHEG: Playability Heuristics for Educational Games PT-BR: Brazilian Portuguese PYA: Practice! Yubimoji AIUEO Q&A: Question and Answer RA: Realidade Aumentada RFID: Radio Frequency Identification RV: Realidade Virtual SRS: Spaced Repetition System TELL: Technology Enhanced Language Learning UI: User Interface ULSJPE: Ubiquitous-Learning System for the Japanese Polite Expressions

Sumário

Lista de figuras ...... ix

Lista de tabelas ...... xi

Lista de abreviaturas e siglas ...... xii

Sumário ...... xiii

1 Introdução...... 16 1.1 Contexto, Problemática e Motivação ...... 16 1.2 Objetivos ...... 17 1.3 Relevância da pesquisa ...... 18 1.4 Estrutura do trabalho ...... 19

2 Estado da Arte ...... 21 2.1 Introdução ...... 21 2.2 O Idioma Japonês ...... 22 2.2.1 Leitura, Escrita, Fala e Compreensão ...... 24 2.3 Estado da Arte ...... 26 2.3.1 Ferramentas para Desktop ...... 27 2.3.2 Ferramentas Web ...... 29 2.3.3 Ferramentas para Dispositivos Móveis ...... 32 2.3.4 Ferramentas Hápticas e de Realidade Virtual ...... 33 2.3.5 Ferramentas de Realidade Aumentada ...... 34 2.4 Desafios ...... 35 2.4.1 Discussão ...... 37 2.5 Conclusão ...... 38 2.5.1 Tomada de Decisão ...... 38

3 State of the Practice of the Web ...... 39 3.1 Introduction ...... 39 3.2 Computer Assisted Language Learning ...... 41 3.3 Methodology ...... 43 3.4 Results ...... 44 3.5 Discussion ...... 51 3.6 Conclusion ...... 52 3.6.1 Decision Making ...... 52

4 State of the Practice of the Android ...... 54 4.1 Introduction ...... 54 4.2 Mobile Assisted Language Learning ...... 55 4.3 Methodological and Theoretical Basis ...... 56 4.4 Results ...... 58 4.5 Discussion ...... 72 4.6 Conclusion ...... 73 4.6.1 Decision Making ...... 74

5 Development of Karuchā Ships Invaders ...... 75 5.1 Introduction ...... 75 5.2 Related Works and Games ...... 77 5.3 Overview of the Japanese Culture ...... 78 5.4 Karuchā Ships Invaders ...... 80 5.4.1 Plot, Gameplay and Features ...... 80 5.4.2 Interface Design and Implementation ...... 85 5.5 Discussion ...... 89 5.6 Conclusion ...... 90 5.6.1 Decision Making ...... 91

6 Interface Evaluation of Karuchā Ships Invaders ...... 92 6.1 Introduction ...... 92 6.2 Karuchā Ships Invaders CALL Game ...... 94 6.3 Interface Evaluation of Educational Computer Games ...... 96 6.4 Karuchā Ships Invaders Study ...... 99 6.4.1 Methodology ...... 99 6.4.2 Results ...... 102 6.5 Discussion ...... 104 6.6 Conclusion ...... 105 6.6.1 Decision Making ...... 106

7 Evaluation of Motivational Aspects ...... 107 7.1 Introduction ...... 107 7.2 Motivation ...... 108 7.3 Method for Evaluation of Motivational Aspects ...... 109 7.3.1 Example of Application of the Method ...... 113 7.4 Discussion ...... 115 7.5 Conclusion ...... 116 7.5.1 Decision Making ...... 116

8 Applying Evaluation of Motivational Aspects on Karuchā Ships Invaders ...... 117 8.1 Introduction ...... 117 8.2 Research and Methodology Scenario ...... 118 8.3 Results ...... 120 8.3.1 Quantitative Results ...... 120 8.3.2 Qualitative Results ...... 123 8.3.3 New Features ...... 125 8.4 Discussion ...... 126 8.5 Conclusion ...... 127

8.5.1 Decision Making ...... 128

9 Conclusões ...... 129 9.1 Trabalho Realizado ...... 129 9.2 Contribuições ...... 130 9.3 Perspectivas e Trabalhos Futuros ...... 131

Referências bibliográficas ...... 132

Capítulo 1 1 Introdução

O uso de ferramentas computacionais para auxílio de diversas tarefas tem tomado cada vez mais espaço no cotidiano das pessoas. Atualmente, computadores, celulares e a Internet têm sido utilizados para os mais diversos fins, incluindo o educacional. O uso de, por exemplo, sistemas tutores, aplicativos em forma de exercícios e jogos educacionais, tem sido amplamente difundido, em conjunto com diversas formas de tecnologia – como a web, computadores pessoais e dispositivos móveis – para dar suporte ao usuário no aprendizado de uma tarefa ou conteúdo ao qual ele não tem disponibilidade de aprender de forma tradicional (como cursos presenciais, com horários e locais fixos que dificultam o acesso àquela informação). No meio educacional computadorizado, uma vertente que vem se destacando fortemente é a de aprendizado de idiomas através dessas tecnologias. Nesse contexto, surge o termo Technology Enhanced Language Learning (TELL). TELL abrange todos os tipos de auxílio oferecidos pela tecnologia – e.g. sistemas tutores, question & answer, jogos e quizzes – para ensino de idiomas, tendo ramificações como Computer Assisted Language Learning (CALL) e Mobile Assisted Language Learning (MALL), que abrangem, respectivamente, uso de computadores e dispositivos móveis para auxílio ao aprendizado de uma nova linguagem. Nesta pesquisa, o foco é no aprendizado de língua japonesa através de ferramentas computacionais.

1.1 Contexto, Problemática e Motivação

O estudo da língua japonesa por si só já é um desafio, principalmente para estudantes ocidentais – como os brasileiros – que não estão familiarizados com sistemas de escrita complexos como é o japonês. Cursos tradicionais em escolas são escassos no Brasil, além de tomarem muito tempo 17 Capítulo 1. Introdução

para serem concluídos, possuírem aulas espaçadas e exigirem que os alunos cumpram um horário determinado em um local específico. O estudo de um idioma como este, por estudantes brasileiros, requer muito esforço, revisões de conteúdo, treinos reforçados de escrita – por tratar- se de novos alfabetos e símbolos para o estudante aprender – dedicados a cada um dos três alfabetos que a língua possui, escuta e leitura de gramática, que difere em muito da língua portuguesa. Estes são apenas alguns dos desafios com os quais vai se deparar o estudante logo ao iniciar seu trajeto no aprendizado do idioma japonês, e tendo isso em vista, é possível afirmar que novas formas de aprendizado e reforço, que possam ajudar o estudante a estar sempre que possível em contato com o idioma são bem-vindas, se feitas e avaliadas especialmente para tal. Nesse contexto, aparecem então os meios computacionais – formas de acessar informações de estudo fora do contexto de uma aula tradicional, e que ainda podem oferecer recursos adicionais de estudo, como áudio, vídeo, recursos de Internet para auxílio de tradução, além de interação com outros estudantes que podem se ajudar entre si. Já existem, em diversos meios computacionais, ferramentas desenvolvidas com este intuito, tanto no meio acadêmico quanto em meios práticos, que vão desde programas para desktop, web, dispositivos móveis e até hardwares específicos, porém como descrito brevemente a seguir, pode-se observar que cada um desses tipos possui também pontos fracos.

1.2 Objetivos

Em vista da necessidade de estudos sobre as diferentes tecnologias computacionais – observando suas vantagens e desvantagens –, que podem auxiliar no aprendizado de língua japonesa, esta pesquisa tem como objetivo principal:  O desenvolvimento de uma ferramenta computacional para auxílio ao ensino da língua japonesa.

O objetivo principal desta pesquisa se desmembra nos seguintes objetivos específicos:  Identificar os principais pontos fortes e fracos das ferramentas atuais existentes no contexto de CALL para língua japonesa;  Analisar as diversas tecnologias/plataformas atuais voltadas para desenvolvimento dos vários tipos de aplicativos encontrados;  Estudar elementos de ferramentas educacionais que possam ser utilizados no design da nova ferramenta;  Desenvolver o design inicial;  Implementar a ferramenta;  Realizar testes de avaliação de interface e interação; Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 18

 Realizar testes avaliação de aspectos motivacionais da ferramenta.

A ferramenta pretende ser voltada para alunos brasileiros de japonês que podem ou não possuir algum conhecimento básico de escrita – i.e. alfabeto hiragana – e visa abranger principalmente o ensino de hiragana, com foco em leitura/pronúncia dos símbolos, e também conhecimento básico sobre cultura do Japão. Pretende-se que o estudo seja proporcionado através do uso de várias mídias (e.g. imagem, áudio e texto).

1.3 Relevância da pesquisa

Em primeiro lugar, é importante contextualizar o ensino-aprendizado da língua japonesa de forma geral, para poder então ressaltar as questões pertinentes ao desenvolvimento de uma ferramenta voltada para o público brasileiro que almeja aprender esse idioma. A língua japonesa é uma língua difícil de aprender, uma vez que tem três alfabetos principais que são todos diferentes do alfabeto romano usado na maioria das línguas ao redor do mundo. Existem os kanji, que são ideogramas originalmente chineses, importados e adaptados para a língua japonesa. Os kanjis representam ideias, e, além disso, um kanji por si só pode ter várias leituras/pronúncias. Os outros dois alfabetos japoneses são hiragana e katakana – que juntos são chamados de kana –, eles são diferentes dos kanji por serem alfabetos silábicos; o hiragana é o mais básico, usado para escrever palavras que não possuem um kanji, como também para os adjetivos e terminações verbais, e como uma legenda – furigana – para kanjis raros ou desconhecidos. O hiragana é o primeiro alfabeto ensinado a qualquer pessoa que comece a aprender língua japonesa, dentro ou fora do Japão. O katakana é semelhante ao hiragana em sua formação básica, contudo os símbolos são todos diferentes; ele é utilizado para adaptar as palavras estrangeiras ao idioma japonês, e em alguns casos para destacar ou enfatizar alguma palavra ou onomatopeia. No Japão, eles também utilizam o alfabeto latino – que é utilizado na maioria das linguagens –, mas muito pouco, como por exemplo, em siglas. A partir dessa contextualização, é possível observar a complexidade do ensino de japonês fora do Japão, e trazendo para o cenário brasileiro, temos, além da dificuldade intrínseca ao idioma, cursos presenciais escassos e caros, e em várias partes do país, até mesmo inexistentes; ao passo que cada vez mais pessoas estão se interessando pelo aprendizado desse idioma, porém não encontram facilmente disponíveis materiais que possam ajudá-los, e que sejam voltados para o público brasileiro, ou seja, material em português-japonês. Em paralelo, temos o enorme crescimento no uso de tecnologias para a maioria da população, como os computadores e a Internet, que proporcionam uma nova janela de conhecimento disponível para a população, e isso inclui materiais de estudo para diversos conteúdos que antes disso não seriam de tão fácil acesso. 19 Capítulo 1. Introdução

Todavia, mesmo com esse crescimento, materiais para o estudo de japonês voltado para brasileiros continua escasso, principalmente em se tratando de aplicativos computacionais, isto é, programas, apps e sites web. Esta pesquisa visa disponibilizar esse conhecimento de forma prática e de fácil acesso, e além disso, apresentando elementos de educação atrelados ao meio computacional, e ainda, em uma forma de entretenimento, através do desenvolvimento de um aplicativo que abranja todas essas características: uma ferramenta computacional de fácil acesso, para ensino básico de japonês, voltada para o público brasileiro, que proporcione aprendizado e diversão ao mesmo tempo.

1.4 Estrutura do trabalho

Neste ponto, é descrito brevemente o conteúdo de cada capítulo apresentado a seguir.

 Capítulo 2 – Estado da Arte: Este capítulo apresenta o levantamento do estado da arte do tema, descrevendo as ferramentas computacionais para auxílio ao ensino/aprendizado da língua japonesa encontradas no meio acadêmico. Além disso, são expostos desafios com relação a desenvolvimentos futuros pertinentes ao contexto apresentado;  Capítulo 3 – State of the Practice of the Web: Este capítulo apresenta os resultados do levantamento de ferramentas CALL Web, voltadas para o contexto de aprendizado da língua japonesa; as quais são também amplamente difundidas através da Internet e utilizadas por estudantes estrangeiros de japonês em várias partes do mundo. Além disso, é apresentado um estudo sobre os tipos de mídia utilizados por estas ferramentas, visando avaliar o comportamento dessas diversas mídias quando a ferramentas são utilizadas em dispositivos móveis com sistema operacional Android;  Capítulo 4 – State of the Practice of the Android: Este capítulo apresenta a identificação de diversos aplicativos MALL (apps) para o ensino de japonês, desenvolvidos para Android, como também a avaliação de seus designs e características de interação. Com base nestes resultados, foi realizado um levantamento de requisitos utilizando O Framework Semiótico, que consiste de um artefato da Semiótica Organizacional, que nos permitiu identificar e esclarecer os requisitos dos apps sob diferentes perspectivas;  Capítulo 5 – Development of Karuchā Ships Invaders: Este capítulo apresenta um estudo sobre a concepção e desenvolvimento do jogo Karuchā Ships Invaders, que tem o intuito de ensinar língua japonesa para alunos brasileiros do idioma, levando em consideração as questões culturais de ambos os países – i.e. Japão e Brasil – e seu impacto sobre os recursos de interface/interação do jogo; Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 20

 Capítulo 6 – Interface Evaluation of Karuchā Ships Invaders: Este capítulo apresenta uma revisão da literatura sobre avaliação de jogos educativos que analisam diversas técnicas, e realiza uma avaliação baseada em um deles, aplicado para avaliação da interface do Karuchā Ships Invaders. A avaliação realizada utiliza uma metodologia baseada em heurística, focada em jogos educativos, e os resultados demonstram a eficácia deste método de aprendizagem de línguas suportadas por um jogo computacional. A descrição do método serve também de guia para futuros avaliadores escolherem métodos de avaliação com base em seu próprio contexto e necessidades.  Capítulo 7 – Evaluation of Motivation Aspects: Este capítulo propõe um método baseado em heurística para avaliar o poder motivacional de jogos educacionais sobre a vontade de um aluno em continuar seus estudos sobre um determinado assunto. O método também é validado com um estudo de caso, cujo cenário é o jogo Karuchā Ships Invaders;  Capítulo 8 – Applying Evaluation of Motivational Aspects on Karuchā Ships Invaders Game: Este capítulo apresenta o resultado da avaliação dos aspectos motivacionais de Karuchā Ships Invaders, através do método de avaliação de aspectos motivacionais – desenvolvido também no decorrer desta pesquisa – demonstrando que esses aspectos influenciam positivamente na vontade do usuário para continuar o aprendizado e a prática da língua japonesa. O estudo foi aplicado com alunos e aspirantes ao estudo do japonês, e os resultados mostram que Karuchā Ships Invaders é uma boa ferramenta CALL para a aprendizagem de língua japonesa;  Capítulo 9 – Considerações Finais: Este capítulo resume os resultados da pesquisa e suas contribuições.

Capítulo 2 2 Estado da Arte1

2.1 Introdução

As relações entre países, seus habitantes e culturas têm se tornado cada vez mais comuns, até mesmo entre nações distantes geograficamente. No contexto da América Latina, a imigração japonesa já não é novidade; japoneses e seus descendentes formam comunidades significativas em países como Brasil, Peru, Argentina, Paraguai e Bolívia. O relacionamento diplomático entre Japão e América do Sul iniciou-se em 1873, com a assinatura de um tratado com o Peru e, em 1895, Japão e Brasil estabeleceram o Tratado de Amizade, de Comércio e de Navegação abrindo, desse modo, o caminho para a imigração japonesa ao país sul-americano. Até hoje, há um expressivo número de imigrantes japoneses na América do Sul, abrangendo desde a primeira geração de imigrantes até descendentes que chegam à quinta geração. O Brasil, por exemplo, possui o maior contingente de japoneses – aproximadamente, 1,4 milhões – fora do Japão. Por sua vez, o início da consolidação da presença brasileira no Japão foi somente a partir de meados dos anos 1990. Brasileiros residentes no Japão geralmente são descendentes diretos de japoneses – também conhecidos como nikkeis – ou casados com descendentes, e buscam no país oportunidades de trabalho e melhoria de vida, muitas vezes mudando-se com toda a família para a Terra do Sol Nascente. Apesar de existir toda essa movimentação de pessoas entre o Japão e o continente latino-americano, existe uma grande barreira: o idioma. Essa barreira é nitidamente percebida, principalmente, para os que se deslocam do ocidente para o país oriental, pois sua língua possui particularidades e complexidades às quais os ocidentais não estão acostumados. Essas características sobressaem sobretudo na forma de escrita do idioma que é bem diferente,

1 Este capítulo é uma versão revisada do trabalho [35], já publicado e apresentado (© 2012 IEEE). Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 22

por exemplo, das línguas oficiais da maioria dos países latino-americanos, ou seja, português, espanhol e francês. Temos observado um aumento significativo na quantidade de projetos de pesquisa e desenvolvimento (P&D) com foco na criação de soluções computacionais que minimizem as dificuldades do aprendizado pleno deste idioma. As soluções computacionais de uma forma geral apresentam um manancial de ferramentas multimídias com a intenção de estimular no aluno de japonês diferentes competências/habilidades exigidas para o aprendizado de um novo idioma. Outra preocupação é a interface de usuário da solução que, geralmente, é contextualizado ao público-alvo. Além disso, também constatamos a existência de diversos trabalhos que apontam a importância, a eficiência, a aceitação, bem como, as vantagens de utilizar o Aprendizado de Línguas Assistido por Computador ou, em inglês, Computer Assisted Language Learning (CALL) como, por exemplo, Lam et al. [20], Tiwari et al. [6], McLoughlin et al. [2], Han et al. [5], Kremenska [4], e Joseph e Uther [3]. Nesse contexto, este capítulo apresenta o estado da arte que descreve os trabalhos da literatura onde ferramentas computacionais são empregadas tendo em vista o subsídio de estrangeiros na aprendizagem do idioma japonês. Além disso, com o intuito de agregar valor a este estudo, identificamos e discutimos diversos desafios computacionais acerca deste tema para a comunidade latino-americana, em especial, desafios relativos à área de Interação Humano- Computador (IHC). Este capítulo está organizado da seguinte maneira: a Seção 2.2 descreve importantes informações acerca do idioma japonês; a Seção 2.3 apresenta o estado da arte; a Seção 2.4 descreve e discute vários desafios computacionais acerca do tema sob a ótica da comunidade latino-americana; e a Seção 2.5 tece as considerações finais.

2.2 O Idioma Japonês

A Língua Japonesa utiliza três diferentes sistemas de escrita: kanji, kana e roma-ji. Os kanji são ideogramas, ou seja, caracteres complexos originados de imagens concretas, que ao invés de representarem sons/pronúncias específicos, representam ideias. Os kanji foram importados da China e adaptados para o idioma japonês. Vários deles são pictográficos, isto é, suas formas lembram o seu significado. A Fig. 1 apresenta um exemplo de kanji pictográfico, que significa “montanha”.

Fig. 1 - Origem pictográfica do kanji de “montanha”. 23 Capítulo 2. Estado da Arte

As pronúncias dos kanji se chamam “leituras”, e cada kanji possui duas formas de leitura, a ondoku – chamada também de on-yomi – que é de origem chinesa e a kundoku – chamada também de kun-yomi – que foi vinculada ao kanji após sua chegada ao Japão. Cada kanji pode possuir diversas pronúncias on-yomi e kun-yomi, aumentando ainda mais o grau de complexidade. Os kanji são compostos por traços de vários tipos, e quando escritos à mão, deve-se seguir uma determinada ordem dos traços, para não gerar ambiguidade na leitura, visto que existem milhares de kanji, e pode acontecer de um ser confundido com outro muito parecido. Kanji mais simples podem ser transformados em radicais, que juntos a outros, podem formar kanji mais complexos, que geralmente representam ideias mais abstratas – verbos e adjetivos – e, consequentemente, com mais traços. Salientamos que essas associações foram feitas quando da chegada dos kanji ao Japão, logo, estão muito relacionadas com costumes da época feudal do país. O exemplo da Fig. 2 ilustra um kanji que “retrata” o fato de os trabalhadores das lavouras encostarem-se às árvores para descansar.

Fig. 2 - Junção de radicais e ideias para formar um novo kanji. Atualmente, o número de kanji que são, obrigatoriamente, ensinados nas escolas – chamados jouyou kanji – é de 2.136 ideogramas; até o ano de 2010 eram 1.945. Isto demonstra que o japonês é um idioma “vivo” que tende a evoluir com o passar do tempo. Já os kana são divididos em duas partes: hiragana e katakana. Cada uma delas é um alfabeto silábico – ou silabário –, onde cada símbolo representa uma única e fixa pronúncia. O hiragana foi desenvolvido a partir da evolução cursiva do kanji e é utilizado para escrever palavras que não possuem kanji correspondente, ou que seu kanji seja muito raro ou desconhecido, bem como para terminações/flexões de verbos e adjetivos. Neste último caso, é chamado de okurigana. Outra função do hiragana é indicar a pronúncia de um determinado kanji, sendo escrito ao seu lado ou sobre ele, formando uma espécie de legenda, chamada furigana. O furigana é utilizado em textos em geral – por exemplo, jornais, revistas e artigos – nos kanji que não pertencem à lista dos jouyou kanji. O katakana também deriva dos kanji e é, atualmente, utilizado principalmente para transcrição/adaptação de palavras estrangeiras em geral, sejam elas substantivos, adjetivos ou nomes próprios, mas também é utilizado quando se quer dar destaque a alguma palavra, mesmo de origem japonesa, e ainda para escrever onomatopeias. Ambos hiragana e katakana possuem 46 símbolos/fonogramas básicos, e alguns deles podem ser modificados com o uso dos acentos japoneses dakuten – mais conhecido como tenten – e handakuten – comumente chamado de maru – que transformam a pronúncia de certos caracteres. Além disso, símbolos também podem Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 24

ser combinados entre si para formar novas pronúncias mais peculiares. A Fig. 3 mostra exemplos de uso de tenten/maru e a combinação de certos caracteres.

Fig. 3 - Uso de tenten, maru, e junção de caracteres. O chamado roma-ji nada mais é do que o alfabeto romano, utilizado na maioria dos idiomas ao redor do mundo, inclusive nas línguas latino-americanas. No Japão é pouco utilizado, sendo visto, especialmente, em indicações em locais públicos e abreviaturas. A Fig. 4 apresenta os termos América do Sul e América Latina em kanji/hiragana/katakana/roma-ji. Em relação a esta figura, não existe kanji para América Latina.

Fig. 4 - América do Sul e América Latina escritos em (a) hiragana (b) kanji (c) roma-ji (d) katakana.

2.2.1 Leitura, Escrita, Fala e Compreensão

O aprendizado íntegro da Língua Japonesa para um estrangeiro exige demasiado esforço, dedicação e tempo. Por isso, muitas vezes esse aprendizado é dividido de acordo com as necessidades de cada pessoa, sendo deixados de lado os fatores considerados menos importantes ou os mais difíceis. O estudo da fluência em fala e compreensão é muitas vezes o alvo principal dos estudantes de japonês, mas para aqueles que pretendem viver no Japão, é essencial que tenham também fluência na leitura e na escrita. Os desafios trazidos pelo estudo da fala e compreensão/escuta são semelhantes àqueles tidos em qualquer outro idioma, todavia, quando se fala na leitura e escrita japonesa, é onde a maioria das pessoas encontra a maior barreira, a qual faz muitos desistirem do estudo deste idioma. Existem até mesmo estudos sobre como o cérebro processa e retém as informações de leitura referente ao japonês. Por exemplo, Mi et al. [1], mostram que há uma diferença de aprendizado, quanto ao armazenamento na memória e as áreas ativadas do cérebro humano, entre kana e kanji. Os resultados desse estudo mostraram que o lado esquerdo do cérebro é ativado primeiro quando é exibido um kana, pois se trata de uma leitura fonética com pronúncia fixa, enquanto que o lado direito é ativado primeiro quando da exibição de um kanji, pois sua 25 Capítulo 2. Estado da Arte

imagem é uma mistura de semantema e morfema2, isto é, representa ideias e não uma pronúncia fixa. Portanto, existe até mesmo a questão de o cérebro dos estrangeiros não estar acostumado a processar a leitura da mesma forma que os japoneses, e esta questão se aplica ao contexto latino- americano, cujos nativos apresentam mais dificuldade em aprender kanji, e sendo estes uma parte fundamental do idioma japonês, isso impõe mais uma barreira ao ensino-aprendizado. Existem diversos pormenores que são fatores influenciadores no resultado do aprendizado desta língua, entre eles podemos destacar o fator cultural de hierarquia e polidez, a compreensão de contextos e situações sobre os quais o idioma é aplicado, não devendo o orador transgredir as regras de cortesia exigidas pelo idioma. No idioma japonês existem também as onomatopeias, que são figuras de linguagem de uso amplo na Língua Japonesa, atreladas a expressões corporais, mímicas, também muito difundidas no idioma. Além disso, muito utilizados no Japão são os provérbios japoneses, chamados de kotowaza. Um método manual tradicional de aprendizado bastante utilizado no estudo da Língua Japonesa é o Sistema de Repetição Espaçada ou, em inglês, Spaced Repetition System (SRS). O SRS utiliza flashcards, que são cartões que contêm questões sobre pequenos tópicos de determinada matéria – por exemplo, um kanji com tradução/leituras, ou uma frase com explicação de elementos gramaticais – para aprender e revisar conteúdos, e à medida que se aprende os assuntos neles impressos, os cartões são classificados de acordo com o nível de facilidade que o aluno teve de lembrar a resposta à pergunta daquele cartão, e é atribuída a cada um a média de tempo – em horas, dias, meses e até anos – que o cérebro do aprendiz levará para esquecer a informação. Esse método possui versões computadorizadas, sendo a mais conhecida o software Anki que, além de outros idiomas, inclui pacotes específicos para o estudo de japonês. Ainda em relação à contextualização deste idioma, vale comentar a existência de um conhecido exame oficial de proficiência da Língua Japonesa, aplicado em várias partes do mundo, denominado Japanese Language Proficiency Test (JLPT). Este teste é dividido em 5 níveis, cada qual com um grau de dificuldade (sendo o 5 o mais fácil), e visa medir e certificar o nível de proficiência de pessoas cujo idioma nativo não é o japonês. O teste é oferecido/realizado duas vezes por ano – contudo em alguns países a frequência é anual –, e qualquer indivíduo cuja língua-mãe não seja o japonês pode prestá-lo. Por fim, gostaríamos de explicitar que nesta seção delineamos os conceitos mais relevantes da Língua Japonesa que foram objetos de pesquisa dos estudos que compõem o estado da arte dos trabalhos que apresentam ferramentas computacionais para apoiar o processo de ensino- aprendizado da Língua Japonesa. Portanto, com essa seção não pretendemos fornecer nenhuma espécie de tutorial inicial do idioma, tampouco ensinar japonês, mas apenas melhor contextualizar o trabalho e organizá-lo de forma clara, visando proporcionar o pleno entendimento deste estudo em toda a sua extensão.

2 Semantema e morfema são termos da linguística que representam, respectivamente, radical de uma palavra onde está seu sentido fundamental, e fragmento mínimo capaz de expressar significado. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 26

2.3 Estado da Arte

O estudo de japonês, especificamente, apoiado por ferramentas computacionais vem sendo cada vez mais empregado. Por exemplo, Kato e Okamoto [7], apresentam as perspectivas tanto de alunos quanto de professores do idioma japonês, sobre o impacto do uso de ferramentas computacionais para o estudo da língua, sendo o estudo voltado para atingir proficiência de nível acadêmico nos estudantes. Os autores desenvolveram um curso com utilização de e-learning composto por quatro módulos de ensino: léxico, gramatical, transferência de informação e módulo contextual. Os dois primeiros são focados em identificar significados de palavras e formações gramaticais, para melhor entendimento de textos acadêmicos; o terceiro módulo é para sintetizar textos e gráficos relacionados, e para isso o usuário deve entender a relação entre os dois, ou seja, como o texto explica o gráfico; e o último módulo é para realizar a interpretação de textos relacionados, entendendo consequentemente a ideia geral do texto estudado. Na validação do estudo, foi aplicado um questionário que visava descobrir: i) quais as diferenças entre as preferências de estudo entre os alunos e os professores; e ii) quais módulos de ensino os grupos consideram importantes para o estudo em um sistema e-learning. Foram entrevistados 22 estudantes estrangeiros de japonês, dentre alunos de universidades públicas e privadas, e 20 professores nativos com experiência variando de 5 a 15 anos de docência. Os resultados mostraram que a maior diferença de opinião entre os dois grupos se encontra no fator “eficácia” de um método e-learning, pois enquanto os alunos demonstraram grande satisfação e facilidade, por sua vez os professores encontraram dificuldades e acharam o método inadequado em certos casos. Outro ponto detectado é que, para os alunos, é mais importante que haja módulos de exercícios tradicionais – vocabulário e gramática –, o que não é tão preferido pelos professores. Contudo, em relação à disposição dos módulos, não houve grande divergência de opiniões. Lee e Chong [8], também atentos a esta tendência, mostram os benefícios trazidos pelo chamado blended learning – pode ser traduzido como estudo misto ou combinado, no qual se mistura o estudo tradicional com o e-learning – no aprendizado da Língua Japonesa. Esses autores utilizaram como base/referência o JLPT, por isso, dividiram o estudo da língua em quatro vertentes: vocabulário de escrita, habilidades de escuta, leitura de gramática e habilidades orais, para então avaliar o estudo tradicional e a eficiência trazida ou não pelo acréscimo do blended learning. No estudo de vocabulário de escrita, os autores frisam a existência de numerosos homófonos na Língua Japonesa e sugerem o uso de fóruns, vídeos legendados e dicionários online com vários atributos para adquirir vocabulário e ajudar a distinguir as palavras de mesma pronúncia. Para o estudo de escuta, é sugerido o uso de gravações em formato MP3, as quais podem estar disponíveis para os alunos estudarem fora do horário da aula presencial, além do uso 27 Capítulo 2. Estado da Arte

também de vídeos com o intuito de aprender a linguagem do dia-a-dia e também dialetos. No estudo da gramática, os autores apontam a dificuldade de estudar as partículas japonesas e indicam o uso de aplicativos que deem feedback para o usuário fixar o conteúdo, importante principalmente para aqueles que irão prestar o JLPT. Para o estudo da oralidade, os autores destacam a importância da hierarquia usada na linguagem no Japão, das formas mais polidas às mais coloquiais, e aponta como uma forma de estudo o uso de aplicativos direcionados a isso, com feedback, mas com a indispensável aula presencial com professores nativos, e uso de áudio para praticar pronúncia. Nas próximas subseções apresentamos as soluções computacionais agrupando as ferramentas da seguinte maneira: a) Desktop; b) Web; c) Dispositivos Móveis; d) Hápticas e de Realidade Virtual (RV); e) Realidade Aumentada (RA). Destacamos que os trabalhos que serão apresentados a seguir foram encontrados nas principais bases de pesquisa científica, isto é, nas bibliotecas digitais da IEEE e da ACM, ISI Web of Knowledge, Scopus e Scielo.

2.3.1 Ferramentas para Desktop

O primeiro trabalho que apresenta um aplicativo para desktop visando auxiliar o ensino/prática de escuta/compreensão para alunos com nível avançado da Língua Japonesa é o trabalho de Mizumachi et al. [26]. O curso, distribuído em CD-ROM, utiliza vídeos onde dois estudantes de japonês conversam de maneira informal, abordando temas tanto sobre a linguagem em si, quanto sobre estratégias de ensino-aprendizado e também sobre a cultura e modo de vida do Japão, fator considerado importante pelos autores, que se basearam em pesquisas específicas. As explicações gerais do CD-ROM estão ainda disponíveis nas línguas japonesa, chinesa, coreana e inglesa, podendo ser selecionadas pelo usuário a qualquer momento. Através de testes de validação, os autores puderam perceber que o sistema teve boa aceitação entre os usuários/estudantes avançados de japonês. Outra ferramenta para desktop é apresentada por Stubbs [28], e é conhecida como o jogo Kana no Senshi. O jogo tem como objetivo ensinar os estudantes iniciantes de japonês a lerem kana – hiragana e katakana – de forma rápida, para poderem iniciar seus estudos de kanji mais eficientemente. O aplicativo é capaz de analisar o desempenho, isto é, a rapidez e a acuidade do jogador e à medida que este progride, os zumbis – inimigos que trazem um kana consigo, são derrotados com o acerto do jogador – se tornam mais rápidos, também para estimulá-lo a ler e identificar os kana mais depressa, tornando assim o jogo possivelmente eficaz no ensino e também atrativo e divertido. Os autores defendem, baseados no processo de validação do seu sistema, que ele é mais bem aceito pelos usuários, pelo fato de atrelar diversão ao aprendizado. A ferramenta apresentada por Namatame et al. [19] de interface intuitiva, voltada a crianças, para o ensino de yubimoji, que é a língua japonesa de sinais. A ferramenta, chamada de Practice! Yubimoji AIUEO (PYA), primeiramente tem a vantagem de possuir apelo em sua Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 28

interface, para que possa motivar iniciantes, principalmente crianças nas escolas, a aprender o alfabeto de sinais básico, diferenciando-se assim dos demais existentes. PYA possui os seguintes recursos: representação visual de caracteres, tanto do lado de quem diz quanto do lado de quem ouve; apresentação, simultaneamente, do kana correspondente, a forma dos dedos, da boca e o som de cada um; animações relativas à forma dos dedos e da boca; animações do movimento dos dedos, e uma explicação sobre a origem do símbolo, servindo como uma dica para memorização. Já Yusuke et al. [9], focam seu trabalho nas dificuldades encontradas pelos alunos na maneira de falar e se expressar, devido à existência dos dialetos, gírias e expressões idiomáticas no idioma japonês. Os autores propõem a criação de um sistema que fornece ensino explicativo e também ensino contextualizado sobre gírias e expressões, para uso da linguagem em diversas situações. O programa desenvolvido possui três funções básicas: reaver expressão, ensinar expressão e o agente gráfico de interação com o usuário. A função de reaver pode ser utilizada de duas maneiras: uma é inserir uma expressão ou partes dela, para que o sistema revele seu significado; a outra é inserir uma situação para que o sistema sugira expressões adequadas a ela; o sistema também fornece exemplos de uso e expressões similares. A função de ensinar se utiliza de animações para facilitar o ensino-aprendizado e também possui a aplicação de testes/provas, e o progresso do usuário é gravado no banco de dados do sistema. O agente gráfico serve para ajudar nas operações do usuário e também são usados vários agentes que conversam entre si ensinando a utilizar expressões. Em [22], Miyazaki e Tomimatsu apresentam uma proposta de design de uma ferramenta educacional para crianças, para ensino de onomatopeias japoneses, utilizando caracteres – hiragana –, sons e gestos, em uma plataforma de interação multi-touch. O sistema, chamado Onomato Planets, é voltado principalmente para crianças da pré-escola, mas também pode ser utilizado por crianças mais velhas e por estrangeiros. Ele exibe caracteres hiragana como “estrelas” e estas podem ser ligadas formando “constelações” que são as onomatopeias, e é dividido em dois segmentos/formas de interação: um para pessoas que não conhecem o hiragana – como alguns estrangeiros e crianças muito pequenas –, onde o usuário pode aproveitar o som e o ritmo das onomatopeias e construir os seus próprios, fazendo ligações entre caracteres hiragana; e outro para aqueles que conhecem o alfabeto, que podem criar onomatopeias através dos gestos e toques na tela. O sistema detecta diferentes tipos de toque através de vibrações, tais como, “tap”, “hit” e “knock”, através do tamanho da área tocada e também do tempo que dura cada toque, e isso gera diversas possibilidades, que são mapeadas através do Onomato Planets. Sa et al. [24] trazem uma ferramenta que auxilia alunos chineses de japonês, que devido a particularidades fonéticas de seu idioma nativo, sentem dificuldade em dominar alguns sons da fonética japonesa, a melhorar/desenvolver sua escuta/compreensão do idioma nipônico, focando-se em sons longos, modificados – uso de tenten e maru – e contraídos, representados no alfabeto katakana. A interface de usuário é composta basicamente por duas funções, a de escuta de exercício e a de resultado de exercício. Internamente, o aplicativo possui as funções de 29 Capítulo 2. Estado da Arte

identificar, analisar e mostrar os pontos fracos do usuário no aprendizado, e então gerar novos conjuntos de exercícios baseando-se nos pontos fracos detectados. O usuário então acessa inicialmente a função de escuta de exercício, na qual ele aprende a pronúncia/leitura de certas palavras, e também responde a testes, e então na função de resultado, são mostrados os acertos e erros, e ao lado dos últimos, explicações detalhadas sobre o erro cometido. A partir daí, o sistema gerará um novo conjunto de perguntas, focando na dificuldade do estudante, assim ele pode aprender as pronúncias através de repetição espaçada. Tiwari et al. [6], apresentam um sistema de tutoria voltado para o contexto de alunos indianos de japonês, por isso, ele é exibido em inglês, híndi e tâmil (as duas últimas estão entre as línguas oficiais da Índia). O sistema conta com os seguintes recursos: banco de dados de flashcards, que apresenta figuras de kanji e uma série de palavras ao lado para que o usuário selecione as relacionadas à figura em questão, exibindo também o número de traços do kanji; ordem dos traços animada, que oferece animações coloridas para cada kanji, ensinando como se deve escrever, além de prover a opção “Speech” para o usuário poder ouvir as pronúncias associadas a cada um; teste e questionário, no qual é dada uma imagem de kanji e o usuário deve escrever seu significado em uma caixa de texto, dispondo também de um botão para dicas, caso o usuário necessite.

2.3.2 Ferramentas Web

Ochi et al. [10] apresentam o desenvolvimento de uma ferramenta para auxílio a estudo de kanji, que utiliza o que os autores chamam de Resources, que são textos japoneses para que os estudantes tenham um estudo contextualizado e aprendam o idioma do dia-a-dia. A ferramenta é capaz de direcionar textos digitais selecionados por usuários para um dicionário e guardar o progresso do usuário, ou seja, os kanji pesquisados e aprendidos. O sistema desenvolvido, chamado JUPITER, possui o papel de filtrar os kanji contidos nos Resources selecionados pelo estudante de nível intermediário e com certa afinidade com kanji através da Internet, para que o ensino esteja de acordo com o nível e as necessidades do usuário em questão. O JUPITER possui os seguintes recursos: inserção automática de furigana; geração de hyperlinks das palavras para um dicionário de kanji; monitoramento de frequência de kanji e detecção do nível atual do usuário; filtro de kanji “furiganizados”, com base no monitoramento, para não inserir furigana nos kanji de leitura já conhecida pelo usuário; criação de um modo practice, baseado nas informações de estudo através dos Resources. O sistema também é capaz de classificar, com cores, os kanji em: kanji que o usuário deveria entender/ler (em cinza); kanji que não estão na base de conhecimento (em vermelho); kanji que estão na base de conhecimento (em azul); e kanji cujo grau de leitura correta é baixo (em verde); e fornece opções para os kanji destacados. Para determinar quais kanji aparecerão com legenda ou não, a cada kanji computado Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 30

é associada uma frequência, e aqueles com mais aparecimento e acertos no modo practice são considerados conhecidos e não mais é mostrado o seu furigana. Em [18], Pershin et al. descrevem uma ferramenta baseada na chamada Dedução Contextual: tentar adivinhar o significado de uma palavra ou expressão desconhecida pelo contexto, buscar informações em dicionários ou listas, ou simplesmente pular a tal palavra/expressão do texto. A ferramenta, que eles chamam de Sistema B, é voltada para ensino de japonês, a partir do inglês, e basicamente faz o seguinte: busca um site com determinado conteúdo em japonês, e outro de conteúdo similar, em inglês, agrupando por data e tópico, então traduz artigos do japonês para o inglês e guarda essas informações em um banco de dados. Quando o usuário navega por uma página que esteja contida no banco de dados, é exibida por um agente virtual a tradução instantaneamente. O Sistema B faz a tradução do texto como um todo, possui vocabulário dinâmico. Contudo, apenas pode ser aplicado a um número limitado de sites. Breen [30] apresenta um dicionário web multi-linguagem que tem o japonês como idioma pivô, cujas entradas possuem tradução para o inglês, alemão e francês. Grifamos que essa ferramenta é uma das poucas que contempla algum idioma falado na América Latina, no caso o francês. É um dicionário simples, que contém organizadas as palavras em japonês – kana e kanji –, pronúncia em kana e suas traduções para o idioma selecionado dentre os oferecidos. Este aplicativo é disponibilizado gratuitamente e inclusive para ser incorporado em softwares de dicionário em geral, como também está disponível em vários sistemas servidores na Internet. Por sua vez, Yusuf et al. [11] aprimoram um sistema anteriormente desenvolvido pelos mesmos autores, chamado Onomatopedia, que é um dicionário online de onomatopeias para estudantes da Língua Japonesa. Eles propõem então o Onomatopeta!3. Os autores frisam o fato de que necessitam da ajuda de falantes nativos de japonês, porém que para que eles ajudem, é preciso que o sistema lhes ofereça algum atrativo. Além disso, os autores desenvolveram dois pequenos jogos na aplicação: o Oshoot, que apresenta para o usuário, quando do upload de uma onopicture, diversas palavras que podem estar relacionadas com sua onomatopeia, para que ele atire e acerte nas palavras que ele pretende expressar através da foto compartilhada, com a opção também de adicionar palavras; e o Slider Part, que mostra, quando o usuário seleciona uma onopicture do site, sentenças similares para que o mesmo avalie-as e classifique-as de acordo com a imagem. Em [21], Zhenlei et al. apresentam uma ferramenta web na qual o usuário tem acesso a um banco de dados de palavras e suas pronúncias em forma de áudio, gravado por nativos japoneses com o sotaque natural do Japão. O objetivo é auxiliar estudantes estrangeiros no Japão a

3 O Onomatopeta! é uma espécie de rede social, na internet, na qual nativos japoneses podem postar fotos com atrelados onomatopeias – nomeadas pelos autores de onopictures – e compartilhar com outros nativos e também estudantes estrangeiros do idioma, e devido ao fato de existirem inúmeros onomatopeias na língua, tanto estudantes japoneses quanto nativos podem se beneficiar e aprender com o serviço. 31 Capítulo 2. Estado da Arte

melhorar seu sotaque ao se comunicar no idioma japonês, visto que para os nativos do Japão é difícil compreender um estrangeiro cujo sotaque ao falar em japonês é afetado por sua língua nativa. O usuário desta solução pode ouvir a pronúncia e gravar ele mesmo falando, fazer upload de sua gravação e esta ser analisada pelo algoritmo, comparando com a original, para ver se está semelhante/correta. Uma avaliação mostrou que o sistema é realmente capaz de analisar as gravações com precisão aceitável. Já Liu et al. [12], desenvolveram um sistema de e-learning baseado em Question & Answer para ensino do idioma japonês. É proposto um sistema online no qual os usuários possam depositar perguntas/dúvidas que surgem após aulas presenciais, e onde os professores tenham acesso e possam respondê-las. O sistema segue basicamente dois princípios: o de atender aos questionamentos dos usuários, tendo quem os responda, e o de não sobrecarregar os professores, criando uma lista de perguntas mais frequentes, para que o sistema seja capaz de respondê-las automaticamente. Os usuários podem agir colaborativamente interagindo com as perguntas dos colegas, e até mesmo os usuários mais avançados que se dispõem a ajudar os mais novatos, podem ser indicados pelos professores para agirem no sistema como uma espécie de monitor, que também tem a liberdade e a função de tirar as dúvidas de outros usuários respondendo seus questionamentos. Outra vantagem é o fato de que, através das perguntas, os professores são capazes de observar e trabalhar nas dificuldades encontradas pelos usuários. Outro sistema web para auxílio no ensino de japonês é o apresentado por Yoshida et al. [13], para ensino de kotowaza, voltado tanto para estudantes estrangeiros quanto para nativos japoneses, que segundo os autores, estão perdendo o interesse em aprender os provérbios nativos. O sistema, chamado JAPES, possui funções de fornecimento de provérbios e seus significados, criação e compartilhamento de material próprio de cada usuário, geração automática de quiz, e é formado por dois módulos: o Proverbial Animation of Ten Seconds (PATS) e o Proverbial Exam by Fourfold Selection (PEFS). No PATS encontram-se as funções de exibição, criação e compartilhamento de animações referentes a provérbios, bem como de compilações de material de cada usuário, as quais podem ser compartilhadas. Já o PEFS é responsável por armazenar informações de aprendizado, aplicar os testes, e mostrar os erros e acertos do usuário. Han et al. [5], que baseado em CALL, apresenta o desenvolvimento de um sistema online para ensino-aprendizado de japonês, no qual são criados cursos e conjuntos de cursos – chamados de “Assuntos” – para os usuários, com oferecimento de exames, e através do feedback desses usuários e de seus dados de aprendizado são retiradas ideias para melhoramentos do modo de ensino. São explanadas as principais funções, que são divididas entre funções para professores – criar novos cursos e assuntos, administrar o aprendizado dos usuários, os cursos e os exames – e funções para alunos que podem escolher de cursos existentes, responder aos testes/exames, opinar sobre as questões, elaborar dúvidas e questionamentos aos professores. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 32

2.3.3 Ferramentas para Dispositivos Móveis

No contexto das ferramentas para dispositivos móveis temos o trabalho de Ichinohe e Suzuki [25], o qual apresentam o aprimoramento de um site web voltado para usuários de celulares chamado The World of Kanji com funções de entrega de exercícios via e-mail, personalização para áreas de maior dificuldade de cada usuário, feedback com o conteúdo correto das respostas para os usuários e anúncio de respostas na própria página web, além de uma maior área de controle para os usuários e um maior número de itens para revisão. Os assuntos fornecidos no site são exibidos para os usuários que prestarão um exame de proficiência em kanji específico no Japão, e mesmo não sendo possível, devido a limitações tecnológicas, prover grandes quantidades de conteúdo, a vantagem de utilizar o conteúdo de forma portável e com flexibilidade de lugar e tempo, compensa essa carência. Outro trabalho interessante é o de Yin et al. [14], que exibem uma ferramenta, baseada em computação ubíqua, que auxilia no ensino do idioma japonês, focando nas expressões da forma polida de conversação. Eles se baseiam nas vantagens da computação ubíqua e distribuída, interação e colaboração entre pessoas. O sistema é baseado em outro desenvolvido anteriormente pelos autores, o Japanese Polite Expressions Learning Assisting System (JAPELAS)4, para uso em PDA. Os autores resolveram expandir as capacidades do JAPELAS, criando um novo sistema, chamado Ubiquitous-Learning System for the Japanese Polite Expressions (ULSJPE), que conta com uso de tecnologia RFID, GPS e wireless LAN, e tem as funções de armazenar os dados de aprendizado – frases em contextos das formas polidas, juntamente com a localização obtida através das tecnologias de rede – do usuário em um servidor, disponibilizando essas informações para outros usuários, para que também possam aprender com as experiências de outros estudantes. Já Lin et al. [15] apresentam um sistema para dispositivos móveis que auxilia estudantes estrangeiros de japonês a aprender kanji por meio de estórias mnemônicas5, em uma interface que utiliza tanto imagem quanto áudio. Como base, os autores utilizam o método apresentado no livro Remembering the Kanji de Heisig, famoso entre os estudantes de japonês, mas também referenciado em pesquisas acadêmicas, agregado ao ensino-aprendizado da pronúncia dos kanji e conhecimento dos componentes dos mesmos.

4 JAPELAS trata-se de um sistema para ensino-aprendizado de expressões japonesas da forma polida, no qual os usuários inserem verbos que desejam utilizar e o sistema mostra quatro formas: casual, básica, formal e muito formal, que correspondem de forma crescente ao nível de polidez da situação. 5 Mnemônica é um auxiliar de memória. São, tipicamente, verbais, e utilizados para memorizar listas ou fórmulas, e baseiam-se em formas simples de memorizar maiores construções, baseados no princípio de que a mente humana tem mais facilidade de memorizar dados quando estes são associados à informação pessoal, espacial ou de caráter relativamente importante, do que dados organizados de forma não sugestiva – para o indivíduo – ou sem significado aparente. Porém, estas sequências têm que fazer algum sentido, ou serão igualmente difíceis de memorizar. 33 Capítulo 2. Estado da Arte

A implementação do sistema foi feita para um dispositivo já existente, um gravador de áudio chamado COWON iAudio X5. O programa exibe uma lista de significados/traduções de um lado, as quais podem estar verdes, se o usuário já aprendeu aquele kanji – criou/gravou uma estória – ou vermelho, se ele ainda não o aprendeu, e a imagem do kanji é mostrada do outro lado, oferecendo recursos como lista de palavras correspondentes às partes que formam aquele kanji, e hyperlinks para estas partes. Ainda é possível utilizar o dispositivo em modo de revisão, no qual ele utiliza flashcards, nas quais é mostrado apenas o significado do kanji, para que o estudante se recorde da sua forma, através de estórias criadas e inseridas no dispositivo pelo próprio usuário. No contexto de dispositivos móveis, outra pesquisa relevante é a de Hou et al. [16], onde é aprimorado um projeto anterior desenvolvido por esses autores, que se trata de uma ferramenta web chamada Japanese Mimicry and Onomatopoeia Learning Assistant System (JAMIOLAS), que utiliza ciência de contexto voltada para o uso de mímica e onomatopeias (MIO) da Língua Japonesa. Os autores se motivam no fato de que o uso de ferramentas computacionais cientes de contexto facilita o aprendizado de palavras de línguas estrangeiras, mas é flexível à necessidade do usuário, principalmente, do ensino do idioma japonês que é muito rico no uso de mímica e onomatopeias em sua comunicação do dia-a-dia. A JAMIOLAS já possuía até então duas versões (1.0 e 2.0), mas foi na versão 3.0, voltada para dispositivos móveis, que os autores melhoraram a função de sensores e detecção em geral do ambiente no qual se encontra o usuário, utilizando através da web informadores de tempo – clima – e localização, além de sensores no dispositivo para detectar outros detalhes. Feito isso, é apresentado para o usuário um quiz com perguntas do tipo “como descrever o tempo agora com o uso de MIO”, além de oferecer outros modos/funções de ensino, com uso de mídias pré- gravadas ou no modo dicionário, onde o usuário pode consultar as MIOs que desejar. O sistema também dispõe de um módulo para professores, no qual estes podem avaliar e validar novas MIOs.

2.3.4 Ferramentas Hápticas e de Realidade Virtual

Solis et al. [29] apresentam uma ferramenta para ensino de escrita de kanji através de uma interface háptica, além do uso de tecnologia Reactive Robots, que é capaz de emular a presença de um tutor humano, sendo neste caso, uma forma de imitar o processo ensino-aprendizagem tradicional de caligrafia de kanji, no qual o professor guia a mão do aluno enquanto este segura o pincel, para ensiná-lo como escrever o kanji, ou seja, ordem dos traços, mais ou menos pressão com o pincel. Desta forma, é possível então automatizar essa vertente do ensino de japonês, não sendo preciso que um professor humano esteja todo o tempo ao lado do aluno. Já Nishino et al. [23] apresentam uma ferramenta para auxílio ao ensino da caligrafia de kanji japoneses, baseando-se na forma de ensino tradicional desenvolvendo a ferramenta com o Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 34

intuito de melhorar as condições de estudo, já que no método tradicional o professor e o aluno precisam estar no mesmo ambiente, além de que o professor só pode atender um aluno por vez, e da forma como ele fica posicionado para escrever – em pé ao lado do aluno –, às vezes os traços não saem fiéis ao natural. O sistema, desenvolvido com RV, é baseado em uma interface háptica, ou seja, que utiliza detecção de toques de forças, vibrações e movimentos. Utilizando um tablet e dispositivo háptico chamado PHANTOM OMNI™, o sistema é capaz de gravar os movimentos do professor-usuário ao escrever um kanji, e repeti-los para que o aluno-usuário os sinta, imitando a forma tradicional de ensino como se o professor de japonês estivesse guiando a mão do aluno.

2.3.5 Ferramentas de Realidade Aumentada

Outro trabalho relevante é o de Wagner e Barakonyi [27], que utiliza tecnologia de RA para ensinar o significado de kanji. O sistema basicamente é feito para PDAs, que são então equipados com detectores – uma câmera – de marcadores de RA além de conexão wireless para interagir com outro PDA, e consiste em um jogo portável para duas pessoas, onde vários marcadores são dispostos em uma mesa. Cada marcador possui dois lados com o mesmo kanji, porém um detectável pelo PDA e outro não, para que os jogadores insiram no dispositivo seu significado sem ajuda. Para conferir as respostas, vira-se o marcador para o lado identificável, e então o PDA irá exibir, em cima do marcador, a imagem em 3D correspondente ao significado daquele kanji, e caso o jogador tenha acertado, marca um ponto. Os autores assim demonstram um aplicativo simples, porém atrativo e educativo a respeito de kanji japoneses. Por sua vez, Kanev et al. [17], visam construir modelos 3D de kanji para ajudar na memorização e entendimento, visto que os kanji, em sua maioria, são representações de objetos/artefatos existentes no mundo real, porém quando escritos se transformam em linhas de duas dimensões. O sistema, chamado 3D Kanji, é baseado em um algoritmo que consegue transformar uma imagem 2D em um artefato 3D, e então o software produzido foi adaptado para, através deste algoritmo, fazer a conversão dos kanji. Além disso, também apresenta um recurso com o uso de RA, no qual marcadores com os kanji em 2D são detectados por uma câmera e mostrados na tela em uma forma 3D, e mais, é possível mostrar dois marcadores diferentes e formar um novo kanji. Após descrever em detalhes esses 23 trabalhos, a Table 1 sinaliza o foco de cada um deles em relação ao apoio do ensino-aprendizado da Língua Japonesa, isto é, se visavam apoiar a sua leitura, escrita, fala ou compreensão.

35 Capítulo 2. Estado da Arte

Table 1 - Foco dos Trabalhos da Literatura.

Trabalho Tecnologia Leitura Escrita Fala Compreensão [26] X [28] X [19] X [9] Desktop X X [22] X [24] X X [6] X X [10] X [18] X [30] X [11] X Web [21] X [12] X X X X [13] X [5] X X X X [25] X [14] X Dispositivos móveis [15] X X [16] X X [29] X Hápticas e de RV [23] X [27] X RA [17] X X

2.4 Desafios

Para o contexto da comunidade/público latino-americana, existem diversos desafios no desenvolvimento de ferramentas computacionais para apoiar o ensino da Língua Japonesa, como descreveremos nesta seção. Podemos dividir os desafios identificados, a partir do estado da arte apresentado na seção anterior, em três grandes áreas: interface/interação, contexto e cultura. Os desafios de interface e interação possuem uma grande amplitude de possibilidades, tais quais: o uso de multimídia com interfaces gráficas, recursos de áudio e vídeo para auxiliar no ensino, e tornar o aplicativo mais interessante, porém sem perder o foco do ensino-aprendizado. O uso de recursos de áudio implica em realizar gravações em várias línguas e dialetos, sendo voltado para o público latino- americano. Outra preocupação está relacionada às questões de resolução, no caso de ferramentas Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 36

para dispositivos móveis, e tamanhos/zoom de fonte, já que sendo esta muito pequena, pode tornar difícil a identificação de kanji, por serem em grande parte, caracteres muito complexos. Aproveitando a tendência atual das tecnologias e a natureza humana que impele o ser humano a procurar relacionar-se, ainda é possível o uso de redes sociais através da Internet para o aprendizado/treinamento, através de jogos/competições a respeito de quem está em níveis mais avançados do idioma, ou de aspectos separados (fala, escrita, compreensão, número de kana/kanji aprendidos). As interfaces e a interação podem até mesmo contar com alguma espécie de inteligência artificial, que perceba os pontos fracos do usuário e organize seus estudos de modo que ele reforce os assuntos nos quais têm mais dificuldade, como também o provimento de feedback imediato, para que o usuário tenha prontamente respostas sobre seu desempenho, e possa melhorar. Em relação ao contexto, podemos apontar desafios como a criação de ferramentas mais contextualizadas para a comunidade latino-americana, visto que a maioria das soluções, como apresentado anteriormente, foi desenvolvida para outros públicos-alvo, normalmente países desenvolvidos onde a tecnologia já é de uso mais cotidiano e abrangente. Alguns desafios em relação às questões culturais são: considerar o desenvolvimento de uma interface que possa abranger a cultura japonesa, porém voltando-se para o público latino-americano ou, até mesmo, especificamente, para cada um dos países do continente sul-americano, cujos costumes são bem diferentes do oriente. Atendimento às particularidades dos diferentes idiomas presentes neste continente, bem como trazer aspectos culturais da comunidade latino-americana quando do design de novas propostas de ferramentas computacionais para apoiar o processo de ensino- aprendizado do idioma japonês. Estas ações podem extinguir dificuldades, todavia dada à complexidade do ensino da Língua Japonesa, é possível que tenhamos diferentes desafios, dependendo do foco: leitura, escrita, fala e compreensão. Isso é corroborado pelo exposto no estado da arte, onde muitos trabalhos focam apenas um aspecto. As exceções, como apontado na Table 1, foram [12] e [5]. Muitos dos desafios são individuais, o que dá a entender que ainda mais difícil é desenvolver aplicativos que abranjam mais de um aspecto, ou todos eles, de forma que o aprendizado ainda seja organizado e eficiente, tanto quanto o de ferramentas direcionadas a um só propósito. Com relação a desafios de IHC, há muito a ser explorado, visto que é uma questão pouco considerada. Observando os trabalhos apresentados, notamos que não há neles uma preocupação explícita com a interface e interação do usuário; um raro exemplo é o trabalho de Maciejewski [31], que descreve a forma de interação entre usuário e aplicativo, em uma ferramenta que provê ajuda para melhorar o conhecimento técnico do idioma japonês por estrangeiros que precisam ter acesso a artigos científicos publicados na língua. Por ser um trabalho muito antigo, diversas questões ficam em aberto, pois ele foi desenvolvido baseado na tecnologia de sua época, e por isso descreve formas de interação muito básicas para hoje em dia. Salientamos, também, que raramente encontramos nos trabalhos 37 Capítulo 2. Estado da Arte

científicos da área comentários em relação a avaliações de acessibilidade ou usabilidade das interfaces de usuário dessas ferramentas. Em nosso entendimento, este fato reforça que as questões de IHC detém pouca atenção durante o design de ferramentas computacionais concebidas para apoiar o aprendizado da Língua Japonesa. Nossa intenção nesta seção não foi apresentar uma lista exaustiva de desafios relacionados ao tema, visto que alguns novos podem surgir conforme os andamentos dos projetos de P&D em curso. Os desafios sinalizados estão associados a potenciais pesquisas sendo, portanto, ponteiros de trabalhos futuros para a comunidade latino-americana de computação.

2.4.1 Discussão

Embora seja um tema de relevância para o público latino-americano, que possui uma relação tão forte e significativa com o Japão, este ainda é pouco explorado por diferentes áreas. A reação comum daqueles que necessitam de ferramental tecnológico é a utilização de ferramentas CALL- based objetivando o ensino-aprendizado de japonês por pessoas proficientes em línguas, como o inglês. Agora, apesar de esta pesquisa mostrar a existência de um grande número de ferramentas para falantes do idioma inglês ou do chinês, a utilização destas ferramentas, em alguns casos, pode agregar novos problemas a uma atividade já complexa que é o aprendizado da Língua Japonesa. Ação de traduzir uma ferramenta parece não ser a solução mais adequada para o contexto latino-americano, pois há questões contextuais e culturais envolvidas, além de desafios específicos em cada área do ensino-aprendizado do idioma (fala, leitura, escrita e compreensão). Por exemplo, na questão da fala/compreensão, a fonética japonesa possui sons semelhantes aos do idioma português, o que já não ocorre com o inglês, que possui pronúncia mais “enrolada”. Já no caso de leitura e escrita, é muito diferente uma ferramenta contextualizada para um público como o chinês, que já está habituado aos kanji, do que para o público latino-americano, que utiliza apenas o alfabeto romano. Podemos, ainda, associar o desenvolvimento de ferramentas CALL-based à implementação específica de aplicativos para dispositivos móveis, visto que é notável a produção e o aprimoramento de artefatos de hardware e de software para dispositivos móveis por todo o mundo, e na América Latina cada vez mais pessoas de várias classes sociais têm tido acesso a esses aparelhos. Esses dispositivos permitem o uso de uma CALL mais personalizada ao estilo de vida de cada usuário, pois com eles podem ser escolhidos, por exemplo, horário e local de estudo de forma ainda mais flexível com relação aos computadores pessoais. Para elaborar um ambiente propício às atividades desse ensino, serão necessários estudos direcionados para transcrever/adaptar um método tradicional (analógico), e torná-lo digital sem que se perca o seu valor educacional. Para isso, entendemos ser necessário estudar o contexto do público-alvo e sua cultura, além das preocupações imperativas ao processo ensino-aprendizado. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 38

O estudo aprofundado do Design da Interação neste contexto é imprescindível para que as interfaces computacionais possam, de fato, apoiar o ensino. Assim, tendo os desafios sinalizados na seção anterior em vista como direções concretas de projetos futuros de P&D nessa área, o idioma é apenas uma barreira, que pode ser transposta por meio da observação e implementação de aspectos relacionados à área de IHC, mas não sem antes conhecer, de fato, o estudante, seu contexto e sua cultura. Quando não existe essa preocupação o atendimento é parcial e a lacuna entre o dito e o feito pode se tornar imensurável.

2.5 Conclusão

Este capítulo tratou do aprendizado da Língua Japonesa apoiado por ferramentas computacionais. Para tal, identificamos e apresentamos o estado da arte das ferramentas computacionais já desenvolvidas que visam auxiliar o processo de ensino-aprendizagem da Língua Japonesa apoiada por ferramentas computacionais. Adicionalmente, visando agregar valor ao estudo, identificamos e discutimos desafios desse tema relacionados, principalmente, a aspectos da área de IHC associados a interfaces de usuários para o público latino-americano. Pelo exposto, acreditamos que novas ferramentas computacionais devam ser desenvolvidas para atender aos cidadãos de nosso continente para a aprendizagem mais significativa desse idioma. Como trabalhos futuros, indicamos um estudo aprofundado do Design da Interação das ferramentas identificadas no estado da arte com o intuito de extrairmos requisitos de design de uma nova ferramenta que apoie o aprendizado da Língua Japonesa para, especificamente, o público Latino-Americano. Buscamos, com isso, prover soluções acessíveis e flexíveis para os latinos que necessitam aprender o idioma nipônico, seja por almejarem mudar-se para o Japão ou por aqueles que pretendem estabelecer relações com pessoas de origem japonesa.

2.5.1 Tomada de Decisão

Através desse estudo, foi possível observar que as ferramentas aqui apresentadas não são facilmente disponibilizadas para que os estudantes de japonês de fato usufruam delas. Decidimos então pela realização de um segundo levantamento, no qual buscaremos as ferramentas que realmente são utilizadas no dia-a-dia pelos diversos perfis de estudantes de japonês. O próximo passo de nossa pesquisa se dá na realização de buscas para encontrar essas respostas, bem como analisar cada ferramenta para perceber quais aspectos e características elas proveem aos usuários, entendendo assim se e porque são largamente usadas.

Chapter 3 3 State of the Practice of the Web6

3.1 Introduction

Japanese language uses three different writing systems, i.e., kanji, kana (hiragana and katakana) and roma-ji. Kanji are Japanese ideograms brought and adapted from China, and the Japanese kanji has two kinds of reading (pronounce), i.e., on-yomi (Chinese reading) and kun-yomi (attached to the kanji when it arrived in Japan) and still, they can have several “on” and “kun” readings. Kana are two syllabic alphabets (syllabaries) auxiliary to the language; hiragana is used to write words whose kanji is rare or does not exist, as for greetings, adjectives, verb endings, and also as furigana (also called yomigana or rubi), that is a little subtitle insert beside a kanji to reveal its reading, usually when the kanji is rare. Katakana, on the other hand, is used to adapt foreign words and names to Japanese idiom, as for emphasis of words. Roma-ji is the Roman alphabet, used in most languages in the world, including Portuguese, but in Japan is barely used (e.g., public location indications and abbreviations). Because of the peculiarities described above, Japanese language learning is, itself, a huge challenge, specially for foreign students that are not conversant to its writing system. On the other side, there is a hardship in access good Japanese school/course, as well as the longstanding courses with far-between classes. Nevertheless, with the improvement of the Internet and the ease of access to “computers” that has occur more and more, students were able to see a new way to review content to support their traditional course, or even study topics of interest by themselves. In this context, language learning (LL) supported by computers and Internet became very popular, using softwares developed specially for this, as utilization of online forums for

6 Este capítulo é uma versão revisada do trabalho [43], já publicado e apresentado. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 40

students all around the world to share knowledge and experiences. This kind of study is called Computer Assisted Language Learning (CALL). CALL studies and their association to Japanese language have already been explored by some authors, as Landreth [33], Matsushita et al. [36] and Librenjak et al. [34]. Despite the importance of these works for the subject-matter, it is also known that outside the academic literature there are several computational tools created with the same purpose too, that is, support LL in general. These tools are, mostly, the resources that are really used by the large public of Japanese language students. So, it is our target point to actually know what are them and what are the features that make them different from the ones found at academy. Still, we see the rise of using mobile devices (e.g., smartphones and tablets), and with it the increasing success of Android OS, which is nowadays the most popular mobile system in the world, in addition to be an open system and also pretty accessible, compared to others in the present market. All these factors do motivate us to accomplish a study about the possibility of using, in Android OS, the CALL Web tools for Japanese most used by the students, evaluating its advantages and drawbacks of this direct portability, from a multimedia viewpoint, being it a very utilized and useful appeal in this kind of educational resource. In a previous work, we identified the State of the Art of the subject-matter by the scientific literature, which presents computational solutions developed to support Japanese LL (Marciano et al. [35]). In a different way, the aim of this work is to identify computational tools focused on support of LL that came not from the academic environment, but from the practical; in the context of our research project, this is called State of the Practice. The State of the Practice tools are usually found in Web and are broadcast by search and disclose from the students themselves, at blogs, sites and virtual communities focused on Japanese LL, and there are even tools developed by the students to help their learning process. The tools are used both by self-taught as by students searching for extra support to the LL. These applications usually offer easy access and interfaces that ease use and learning, making them more popular than the academic ones. It is known that for the realization of a study about CALL for Japanese language, it is fundamental to traverse the related academic solutions, that we have done in Marciano et al. [35]. However, we should not leave aside the tools that are really used by students in practice, namely, the tools that compose the State of the Practice of the subject. In this work, we focus on the study of CALL Web tools. This chapter is organized as follows: Section 3.2 contextualizes Computer Assisted Language Learning; after that, Section 3.3 exposes the methodology used to fulfill the work, and then presents the results; Section 3.4 discusses the results focused on multimedia perspective, and, lastly, Section 3.5 presents the conclusions. 41 Chapter 3. State of the Practice Web

3.2 Computer Assisted Language Learning

Over the last decades, many studies were made about the advantages of using Computer Assisted Language Learning (CALL). Being CALL tools, basically, tools that help and ease LL, supporting traditional learning (blended learning); and offering ways for the students to have autonomy in their study (autodidact). These tools are different by the mechanisms used, the interfaces, resources available, by the fact that they favor some specific skills needed for language proficiency. Now, we present some works and examples of results of CALL application, for an overview about using this kind of resource. Mcloughlin et al. [2], for example, made a study willing to demonstrate the enrichment brought to LL by the use of technology and CALL, and still, present as results the advantages of this approach about traditional learning. To perform the study, the authors selected, between Australian students of French, three groups as follows: i) internal students on campus who follow traditional teaching with classroom taught by teachers; ii) distance learners, who occasionally have classroom; and iii) students who have no personal contact whatsoever with teachers and study entirely by mail, using a method multimedia called French in Action, developed at Yale University. Initial tests were performed in 1992 and 1993, and both showed that the group (iii) apparently in more disadvantage, gave results significantly better than the other two groups, i.e., (i) and (ii), however, these studies did not generate concrete statistical data. Years later, in 1996, was experienced to introduce French in Action in studies of groups (i) and (ii) due to the positive feedback regarding them in previous years, and a new study was carried out in detail this time, generating well analyzable statistical data. Again students study only by mail outperformed the other two groups, but with a negligible difference this time, which suggests that the use of the method/system works well for a variety of contexts, and that the evidence that learning supported by computer brings improvements is strong, and that students who study by this method feel more motivated and better understand the content. Another interesting study is that of Joseph and Uther [3], where they discuss the benefits brought by the use of mobile technologies for learning foreign languages, showing the possibility of using many resources such as images, audio and interaction/cooperation with other students. It also explores the blend and complements the behavioral and constructivist approaches in language teaching. First, the authors discuss the use of visual media, explaining that its use – at the time – is limited due to mobile devices have little storage capacity and processing, however, that won this barrier, are of great help as the humans learn a language accumulating many visual experiences, as well as noise; for then, the paper demonstrate that the audio features can be used for various purposes, such as, guide, broker and pronunciation modeler, comprehension and retention of phonemes, conversation practice and fluency in the language, speaking after the barriers that must be overcome, such as quality of audio input and output and/for use by speech Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 42

recognizers, and the size and format of the files, which should be compatible with the device in question. On the other hand, Kremenska [4] conducted a study on how a course with CALL resources can help in traditional study and motivation of language students, and detect which of its factors may be enemies in order to demotivate them or take the focus of learning based on feedback from students. The author developed an experimental course design, in which there are the steps: i) first to test the level of students; ii) then apply two programs with the same goals and content, but with different approaches - one based on CALL and the other not; iii) collect results on improved knowledge about student motivation; iv) process the collected data; v) analyze the data statistically; and vi) completion report on the motivational role of the programs of study. The author details the steps above, dividing them into three stages, in which expands the explanation of what is being done at each step; at the end of it all, the most important is to detect the points that motivate or discourage students the most. Lam et al. [20] evaluate the level of acceptance and support provided by e-learning systems with the use of blended learning, i.e., a mix between traditional teaching and e-learning. The survey was conducted by evaluating an e-course with Japanese students invited to test it and answer an evaluation questionnaire, with scores of 1 to 4, 4 being the best score. About the program as a whole, and its usefulness for learning, the evaluation showed it had an acceptance “good” (score 3)/“great” (score 4) in 95% of users trials, being 48% good, none of the sections of the application received score 1, i.e., bad. The authors conclude that e-learning systems is an important tool to aid the traditional teaching, making it more efficient and enjoyable for the student, especially when it comes to LL, supported by the evaluation of the system, which showed itself as important from the standpoint of students, reaching more than 90% in general acceptance. Also, they explain that this kind of system can help students with different learning rhythms because of the freedom of programming and repetition of content. The authors emphasize that students feel quite interested by sections containing multimedia elements and online interaction. Han et al. [5], focuses on a general theoretical view on the subject, addressing points as modes of applicability, concept and advantages of using CALL, justifying that the increase in technology in general makes CALL be provided with better resources related to quality and diversity, such as multimedia, Web services, sharing information and learning with people around the world. They also explain that teaching with the use of CALL can be considered gainful when combined with traditional school methods, and also provide a more personalized teaching/learning. Tiwari et al. [6], also shed light on some advantages of CALL, such as: flexibility when preparing lesson plans and courses; individualization/independence of the student who has the freedom to determine when to start and stop times of study; immediate feedback provided by the CALL-based tools, good error analysis, helping both the student and 43 Chapter 3. State of the Practice Web

the teacher to identify the difficulties in learning; practice sessions with the use of multimedia, and which the student can carry as many times as necessary. Seeing the advantages of using CALL, defended and certified by several authors, like the ones referred above, in the next sections we present methodology, results of the performed work focused on identify and evaluate the State of the Practice Tools of the subject: CALL Web Tools to support learning, specifically, of Japanese language. The choose for deepen the study of Web tools is due to the accessibility factor, since most of them is easy to access and do not need do install, being able to be accessed from various places, moreover, most of them offer free content in full.

3.3 Methodology

The case study was performed in two steps: i) identification of Web tools of the State of the Practice; and ii) evaluation of the identified tools with focus in placing the problems that occurred in medias outside the environment they were developed to (PC desktop browser). We underline we do not present in an exhaustive manner the tools available in the market, whereas during the development of this work, we found a really expressive number of these kind of tools; since our focus is not to list them, but in the identification of the most used resources and the possibility of portability to a mobile platform, in this case, Android OS. For the accomplishment of this work, we understand the term “multimedia” according to the definition made by Fluckiger [32]. To this author, multimedia appoints the combination, controlled by computer, of text, graphics, image, video, audio, animations, and any other way whereby the information can be represented, stored, transmitted and processed digitally, in which there is at least one kind of static media – e.g., text, graphics, or image – and one of dynamic media – i.e., video, audio, or animations. Along the chapter it is described if the explored tool is available in Brazilian Portuguese language (PT-BR), since this identification is relevant to our research project context. Still, it is important to clarify that, in this work’s context, “input” are the medias who can be given as an entry by the user; and “output” is the way content is presented to the user by the application. We proceed documenting its main functionalities, access availability to PT-BR, and utilized medias, observed in its original environment – the one that each tool was developed to – and evaluate them in a mobile Android environment, to verify how they behave – if they work completely, partly or do not work – with a view from the identified medias for each one, in other words, we evaluate the behavior of the medias at the proposed environment – if and how they were affected when run in this environment, the handicaps found and possible solutions for them. The specifications of the device used in the tests are: Android OS v4.1 – Jelly Bean (API Level 16), CPU ARM, screen size 7’’, with Adobe Flash Player to Android v.11.0.1.153. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 44

The study to evaluate the behavior changes of the present medias in each tool was performed adopting the five following stages: i) 30 exploratory minutes at the original tool environment; ii) selection of three tasks; iii) execution of the selected tasks in the original environment, i.e., a conventional PC desktop browser; iv) execution of the same tasks in Android environment, i.e., Android’s default browser; and v) comparison of behavior between original and Android environment, under a multimedia perspective.

3.4 Results

In the process of identifying the tools for the tests to be conducted in this study, we found a significant number of CALL Web Tools focused on Japanese. However, for the selection of the tools that would be evaluated in detail here, we consider two main points: i) the most widespread tools among students ii) as well as those that make use of multiple media types. Several tools were left out of this review, but our focus is not to provide an exhaustive list and description of tests, but rather evaluate the possibility and difficulties of using such tools and media in an Android mobile environment. The tools identified were classified according to the verbal language skills they serve – i.e., reading, writing, speaking, and listening – and also the tool availability into Brazilian Portuguese. Still, we show the results of tests conducted in Android environment proposed, with all the tools that have not been developed for this platform, as described below. From now on, we will present the selected tools and reports about its features and also its use in the Android environment. Ajax を使った手書き文字認識 (handwriting recognition using Ajax)7 is a web tool that performs recognition of kanji and kanas from a drawing done by clicking on an area of the site itself. As the user draws, it shows possible kanji by the side. By finding the kanji you want, you can click on it and this is placed in a text box to be copied by the user. Ajax does not recognize kanji/kanas written in the incorrect order of strokes (if this occurs, the suggestion of the application is made over the order and number of strokes than shape drawn by itself). Ajax is focused on the study of writing, and uses the media image to input and text to output. On Android, Ajax did not succeed, because even when you select the area to scan the drawing/character, you cannot make a stroke with click and drag, like it is done on the Web by a PC desktop. A tool similar to Ajax is Handwritten Kanji Recognition (HKR)8, but it does not recognize kanas, only kanji. In HKR, for each designed stroke, a small number is printed on the spot where it began to be drawn, indicating the number of strokes made so far (you can disable this option).

7 http://www.chasen.org/~taku/software/ajax/hwr/. 8 http://kanji.sljfaq.org/draw-canvas.html. 45 Chapter 3. State of the Practice Web

Moreover, HKR has more options for how the input should be recognized, and redirects the user to a page with information of the recognized/selected kanji. The study of writing is the focus of HKR, which uses media image for input and text for output. About the test in the Android environment, the same problem of the tool Ajax was presented. Like Ajax and HKR, the online tool uPal9 allows recognition of kanji through drawings made with the mouse. After detected the kanji, you can select it to turn into text and then use it. In addition this tool provides the option to display information about the kanji, for example, showing “on” and “kun” readings, meaning in English and animated stroke order. It recognizes to some extent, kanji written in the wrong stroke order. uPal focuses on the study of writing and reading, and uses the media text and image to input and output. When you access it from the Android browser, a redirect is performed to a special page for mobile and, unlike his fellow Ajax and HKR, uPal works well on the device, but it is slightly harder to get precision in writing. Another tested tool is Denshi Jisho10, which is an online Japanese dictionary to conduct searches for sentences, words or kanji, in Japanese (kanji, kana, or roma-ji) and in English (to find the equivalent in Japanese). After the search is performed, it is possible to direct for example sentences using the search term, if they are registered, as well as details of the characters of the search. On the characters detail page, you can see “on” and “kun” readings of each character, and the translation of each word into Brazilian Portuguese, French, Spanish and English, though, initially, the page shows up only in English. It is also available several information such as stroke order, with representations of drawings, radical, components, and presence/place in Jouyou Kanji11. Denshi Jisho’s main objective is the study of reading, and it uses the media text to input and, text and image to output. It is possible to use normally the Denshi Jisho in the native Android browser. Jiten12 is a dictionary similar to Denshi Jisho but simpler in detailing regarding the kanji, the more, they share the same concept. Jiten is focused on the study of reading and uses the media text to input, and text and image to output. The test was also successful in Android. One more dictionary is WWWJDIC13, which features, besides the translation of Japanese words into English, links to various information about the word sought, usage examples and audio pronunciation. It is a site focused on learning reading and also covers listening. Use the media text as input and text and audio as output. In the tests, it was possible to access media text normally, but the audio could not be played. Still, as a dictionary type of tool, there is Japanese Names Dictionary (JND)14, which presents the English translation of a searched word in Japanese, and shows representative figures to that word, for stronger fixation. It is focused on the

9 http://www.sp.cis.iwate-u.ac.jp/icampus/u/edic.jsp. 10 http://www.jisho.org. 11 List of kanji characters announced officially by the Japanese Ministry of Education (currently 2,136). 12 http://www.jiten.net. 13 http://www.csse.monash.edu.au/~jwb/cgi-bin/wwwjdic.cgi?1C. 14 http://japanese.reader.bz. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 46

teaching of reading through text media for input and, text and image for output. Its behavior was not impaired on the Android device. The Furiganizer15 is a Web application in which a text can be placed in kanji, and it applies through a command, the furigana in the text provided. This application is focused on the study of reading, and media text is used for input and output. The test on Android was satisfactory, indicating that it is possible to use The Furiganizer in the browser of Android environment normally. While the tools RealKana and RealKanji16 are Web tools to perform reading practice of, respectively, kanas and kanjis. The user can choose which set of letters he wants to practice, and the fonts on which they are presented (between different font styles: conventional or simulating handwritten). In practice tab, it appears a certain symbol (RealKana) or word (RealKanji) and the user must enter its reading in roma-ji correctly, then the application proceeds to another symbol, if the answer is correct, or remains the same, showing a ’x’ indicating an error; the user can see the answer by hovering over the kana/kanji. Both applications aims at the study of reading, and use the media: text for input and text and image for output. RealKana site is available in PT-BR. The behavior of RealKana on test environment was not successful, except when using an external keyboard, and not the mobile device touch. RealKanji have not worked with any of the options, external keyboard or touch. The Ultra Handy Japanese Verb Conjugator (JVC)17 allows the user to type, in roma-ji, a verb in its infinitive (dictionary) form, and then returns the conjugated verb in: present indicative, presumptive, imperative, past indicative, past presumptive, provisional and conditional, with an informal and polite form for each one. Being a Web page, it works well in Android browser. Analogous to JVC, there is Japanese Verb Deconjugator (JVD)18, however working in reverse, showing the infinitive form of the verb, as well as seeking to indicate what time/mode of the verb typed. Still, JVD also works with adjectives. JVC and JVD are not in the same page, and are not from the same developer. The JVD also works normally in the browser of Android device. Both are focused on the study of reading – grammar – and only use the media text for input and output. Nihongo Dekimasu19 is a free online course in Japanese, also available in Portuguese, with several multimedia features. Lessons are taught through videos that depict the day-to-day life of a foreign student in Japan, with the possibility to show subtitles in Japanese with kanji, Japanese with kana only, roma-ji, English and Portuguese. In the system, the dialogs can be displayed as text or manga, and there are review questions with immediate feedback. For registered users – for free – it saves a record of study, containing the student’s performance. Nihongo Dekimasu is to study reading and listening, and for this, it makes use of media text, image, audio and video for

15 http://www.furiganizer.com. 16 http://www.realkana.com. 17 http://www.japaneseverbconjugator.com. 18 http://kanjidict.stc.cx/japverb3.php. 19 https://www.erin.ne.jp/en/. 47 Chapter 3. State of the Practice Web

output. In tests, the media have been affected to be used on Android, because despite running the homepage of the site, videos, images and exercises could not be displayed. Moreover, the course of NHK World Radio20, on Android, media image and text were presented correctly but the audio could not be accessed. This course is offered online, where it is also possible that students get the lessons for their devices in audio and text. It focuses on the teaching of reading and listening, using the media text, image and audio for output. And the course Vamos Ganbatear21, offered weekly by Radio Nikkei, offers its audio and text of each lesson for download, also in PT-BR, aiming to teach reading and listening through media text and audio for output. For media text, tests were acceptable, but the audio could not be played. Livemocha22 is a social network focused on LL, where the “teachers” are the users of the network, they teach their native language and in return, receive lessons from natives of the target language of their study. It is possible to send an audio of reading, done by students in the target language for natives to comment and correct. In addition, the site also offers ready lessons in various forms of media, with exercises and feedback. Japanese is included as one of the languages offered by the site, which is focused on the study of reading and listening, as well as speech, corrected by the natives. As used media they have audio for input, audio and text for output. In the tests, the site only showed content with options without Flash, even with the plugin installed. Text and image medias remained the same, but the audio could not be used due to this. With similar methodology, but with a different target audience, Lang-823 is a site of language teaching, where each user makes posts in the language he wants to learn, so it is more suitable for advanced users. People native to the language that the student wants to learn correct your text, your post, and comment on everything in their native language so that the user is actually immersed. The focus of Lang-8 reading and writing with the media text for input and output. The use was not impaired in Android, with the site’s features remaining operative. We also evaluated LingQ24 site, that offers lessons in audio with text to follow, in which the registered user can create a list of known words and their translations, as well as doing exercises with content already learned. It focuses on reading and listening, and uses the media audio and text for output. On Android environment, the tasks could not be performed because the site does not fit the device. However, for users who have a paying account at any level, it is available the LingQ Android App. And JapanesePod10125, also, is a website with a proposal to study Japanese with a podcast, where it is available various multimedia resources for learning, as audio lessons, text to follow, as well as videos and images. Some of the features of the site are paid, but the newly registered user is entitled to a seven day test of all system features. On the

20 http://www.nhk.or.jp/lesson/english/index.html. 21 http://www.radionikkey.jp/por/community/japanese/. 22 http://www.livemocha.com. 23 http://www.lang-8.com. 24 http://www.lingq.com. 25 http://www.japanesepod101.com. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 48

Android device, we could not hear the audio or watch the videos, but the medias image and text remained unchanged. The browser extensions Rikaichan26 – Firefox – e Rikaikun27 – Chrome – are meant to reveal a pop-up when the user positions the cursor over a kanji (text) present on the page, to show information such as translation and some forms of use. Reading is the focus, using the media text for output. For Android browser, there is Rikai, a paid App with the same functionality as the two mentioned above. While Furigana Injector, which is also a browser extension – Firefox28 and Chrome29 –, inserts furigana on the kanji in Japanese texts of a certain tab. This is a complement to the study of reading and uses the media text as input and output. PopKanji30 is a small popup program written in Javascript to run in browsers. With it enabled, the user selects a kanji, and the application shows a figure with its stroke order, as well as its reading and translation, among other information. It is available the translation into PT-BR. It is an application focused on the study of reading, and employs the media text as input, and text and image for output. The application PopKanji did not work in Android browser, because it did not offer the option of selecting uneditable text. Kanji Drill31 is a site that offers an online system of flashcards to registered users. It focuses on the study of reading and uses the media text for input and output. Like it, there is SpeedAnki32, that has its flashcards focused specifically on the JLPT exam33, and uses the media text to output. Both have a role in reviewing the studied cards, as a usual Spaced Repetition System (SRS). Kanji Drill and SpeedAnki work seamlessly on Android, just like in its original environment. Njuku34, which is a site, offers great Japanese texts with audio for accompaniment, and some words translated into English. You can download the audio files. It is focused on the study of reading and comprehension, using the medias text and audio as output. The behavior of the media text in Android was normal, but the audio could not be accessed. Syvum35, e-learning site available in PT-BR, offers quizzes in Japanese for the study of reading and listening through the media text for input, and audio and text for output. By accessing Syvum in the Android environment, the functions were kept, although they are presented differently. However, with the audio, we have not succeeded.

26 https://addons.mozilla.org/en-us/firefox/addon/rikaichan/. 27 https://chrome.google.com/webstore/detail/rikaikun/jipdnfibhldikgcjhfnomkfpcebammhp. 28 https://addons.mozilla.org/en-US/firefox/addon/furigana-injector/. 29 https://chrome.google.com/webstore/detail/furigana-injector/cbahnmcliajmanjkaolemjelphicnein. 30 http://www.aprendendojapones.com/2009/07/30/popkanji-dicionario-de-kanji-em-popup/. 31 http://www.kanjidrill.com. 32 http://www.speedanki.com. 33 The Japanese Language Proficiency Test (JLPT) is a test that evaluates and certifies Japanese language proficiency for non-native speakers. 34 http://www.njuku.com. 35 http://br.syvum.com. 49 Chapter 3. State of the Practice Web

Teaching Japanese for English speakers is the purpose of the site Nihongo o Narau36, taking advantage of explanation lessons with vocabulary, grammar, writing, reading, speaking and listening, and contextualization with Japan using music and photos. The media text, images and audio are used for output. The behavior of all media in Android remained unchanged. Nihongo De Care-Navi37 is a site that offers phrases in Japanese with English translation, with examples of usage and word searches, along with audio for words and phrases. The teaching of reading and listening through the media text for input and, text and audio as output, is its goal. In tests, it offered a version for mobile devices, and all media work correctly. The Anki38 is a desktop program (however on their website is offered access to AnkiWeb, which is a version of the program that can be accessed on the web) of SRS digitalization, which contains specific plugins for studying Japanese. Through it you can create decks of flashcards with the fields “expression”, where you can put kanas, kanji, phrases, “meaning”, for translation of the content of the previous field; “reading” that contains the reading of content, in hiragana alphabet, and “tags” for better organization of content. It is also possible to attach sounds to flashcards, to study listening. Anki is widely used by Japanese students from around the world, and because of this, it is easy to find ready decks, which students can use, as well as create your own decks using them as a way of fixing content learned by other means. Anki can be used for training both reading and comprehension, because of its multimedia features, and the desktop version is available in PT-BR. The AnkiWeb version, which can be accessed from any browser, works fine on Android. Table 2 summarizes 30 identified and evaluated tools, classifies them according to the language skills they support, and also indicates the used medias, i.e., (T)ext, (P)icture, (A)udio, and (V)ideo. Table 2 - Medias used by CALL Web Tools for the Japanese language, identified by language skills. Language skills Available in Tool Reading Writing Speak Listening PT-BR Ajax - T & P - - No HKR - T & P - - No uPal - T & P - - No Denshi Jisho T & P - - - Partly Jiten T & P - - - No WWWJDIC T - - A No JND T & P - - - No Furiganizer T - - - No RealKana T & P - - - Yes RealKanji T & P - - - No JVC T - - - No JVD T - - - No Nihongo Dekimasu T & P - - A & V Yes Curso NHK T & P - - P & A Yes Vamos Ganbatear T - - A Yes Livemocha T - T & A A Yes

36 http://www.learn-japanese.info. 37 http://eng.nihongodecarenavi.jp/eng/index.php. 38 http://www.ankisrs.net. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 50

Lang-8 T T - - No LingQ T - - A Yes JapanesePod101 T & P - - A & V No Rikaichan T - - - No Rikaikun T - - - No Furigana Injector T - - - No PopKanji T & P - - - Yes Kanji Drill T - - - No SpeedAnki T - - - No Njuku T & A - - T & A No Syvum T - - T & A Yes Nihongo o Narau T & P T T & A T & A No Nihongo De Care-Navi T - - T & A No Anki/AnkiWeb T - - T & A Yes

After the identification of different tools, we selected the 26 most used in Practice, presented above, to evaluation by tests in Android environment. In these tests, some tools worked perfectly, others had modifications that did not compromise its performance, while some of them had part of its content spoilt, inaccessible or simply did not work; on the other hand, some of them offered a version specific for the device/environment. The result of behavior of every multimedia element of these tools during the tests in Android environment, as a highlight of what could be lost of the application when used outside its original environment, are summarized on Table 2. We also highlight that, in this table, four of the tools described on Table 1 are not present, for they are plugins for a different browser that the Android default one, being this our default test environment for all of the identified tools. Table 3 joins the results related to the tools that were able to be used somehow in Android environment, showing if there were behavior changes of Text, Picture, Audio and Video medias. Because of that, not all of the tools described above are contained on this table. In its cells, ‘Y’ is positive for change of behavior, ‘N’ is no change of behavior to (I)nput or (O)utput of information in the referred media; and “-” means the media was not used in that tool. Table 3 - State of the Practice Tools, synthesizing media behavior when accessed from the Android platform. Multimedia behaviour in Android OS Tool Text Picture Audio Video Ajax N (O) Y (I) - - HKR N (O) Y (I) - - uPal N (I/O) N (I/O) - - Denshi Jisho N (I/O) N (O) - - Jiten N (I/O) N (O) - - WWWJDIC N (I/O) - Y (O) - JND N (I/O) N (O) - - Furiganizer N (I/O) - - - RealKana Y (I) & N (O) N (O) - - RealKanji Y (I) & N (O) N (O) - - JVC N (I/O) - - - JVD N (I/O) - - - Nihongo Dekimasu N (O) Y (O) Y (O) Y (O) Curso NHK N (O) N (O) Y (O) - Vamos Ganbatear N (O) - Y (O) - Livemocha N (I/O) - Y (I) - Lang-8 N (I/O) - - - LingQ Y (O) - Y (O) - 51 Chapter 3. State of the Practice Web

JapanesePod101 N (O) N (O) Y (O) Y (O) Kanji Drill N (O) - - - SpeedAnki N (O) - - - Njuku N (O) - Y (O) - Syvum N (I/O) - Y (O) Nihongo o Narau N (O) N (O) N (O) - Nihongo De Care-Navi N (I/O) - N (O) - AnkiWeb N (O) - N (O) -

3.5 Discussion

In our analysis, we observed the tools hold manifold focus, medias and strategies for the Japanese LL/support, and it’s up to the interested user to make a choice for a most appropriate tool for his context, even though many of them still cannot be used in mobile environment, ignoring this market trend. Between the identified tools, some of them support students native in different languages, including specificities of the PT-BR. We highlight that, with this work, we presented an overview of the area, containing a considerable number and diversity of content, showing, so, the possibilities and trends of the nowadays State of the Practice, but not an exhaustive list of tools. The tests with tools were made manually, for each one of them; first by the fact that it was necessary to know each tool, the tasks each one perform and the medias they use, since they are very distinct from each other; also, because in the context of the evaluation of media behavior and how it affects the learning process, we chose to do a thorough evaluation, under a human viewpoint instead of automated. A handicap found during the mobile Android environment tests was the Flash Player point, since the platform does not offer support anymore, even is no longer possible to install Flash Player from Google Play®, neither from the Flash Player developer’s official site. To accomplish the tests, it was necessary to search for the player in .apk format, and install it manually, but even like this, the results were not satisfactory. That occurs because of the fact that Flash technology is being replaced by sites developed with HTML5 technology. However, as many sites still use Flash technology, the deactivation of support creates a barrier with the sites that still did not migrate to new technologies, currently supported. About the used medias – i.e., text, picture, audio, and video – by the evaluated tools, it was possible to notice a trend for massive utilization of text and picture medias, followed by audio media, and just a few applications use video media. Observing the Android environment tests, the less affected media was text, as for output there was no changing in any of the applications, whilst audio was the one who suffered more changes of behavior, mostly being not able to be accessed. This can be partly explained by the differences between the artifacts available in each platform of hardware/software, but maybe not by a lack of interest or investment to adapt it, what in some situations does not involve large complexity. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 52

About the language skills supported, there is a remarkable trend focused on reading learning, followed by listening; this can occur because of the complexity required to test abilities like writing or speak, related to its input medias and the way to analyze them. About the tools that are or present version in PT-BR, it is possible to highlight that just a small part of them contemplates native PT-BR speakers public. After made all the tests, we could also observe that most of the tools can be used at mobile environment, i.e., are compatible, however, because they were not developed for that, it is not rare that the user cannot access the tool’s whole potential, mainly over questions about interaction/usability. This could be perceived because the tests were made manually. Through the development of this work, still was possible to observe that, even there are many options available of CALL Web Tools for Japanese, there is still a lot to be done to obtain satisfactory adaptation, according to the new technological reality that are the mobile devices. This warns how much current CALL systems are not prepared to the rise of mobile devices, specifically about Android platform, and its use in LL.

3.6 Conclusion

In this work, we identified the State of the Practice of CALL Web tools that support Japanese LL, and then, we evaluated how these tools behave in Android OS about its portability, media recognition, technologies and strategies contained in them, and educational media. With the achievement of this work, we were able to find that there are several tools for Japanese LL, used by students all around the world. We also performed a case study about the possibility of using these tools in an Android OS environment, and the impact it could cause on the use, regarding the multiple medias used in the process of learning Japanese. We considered the behavior of each media, being used every single evaluated tool outside of its original environment (PC desktop browser). Still, we discussed about the results found in the tools’ tests. As future work, we point the study of tools focused specifically to the mobile Apps. As well as a study about requirements and the development of a very own tool, that embrace the real nowadays public of Japanese language students, that use digital environments offered by the new technologies with educational purpose.

3.6.1 Decision Making

After the accomplishment of this survey, it was revealed that there is a great tendency for the various applications to migrate to the context of mobile devices. Being Android the most widely 53 Chapter 3. State of the Practice Web

used system in the world today in such devices, we decided by the search tools focused on teaching Japanese that are available in this context. Through this survey, we also intend to highlight more features and requirements that such a tool usually has, intending to start a project to develop a new tool, with a focus that is little embraced by the existing tools, and addressing new research prospects.

Chapter 4 4 State of the Practice of the Android39

4.1 Introduction

Japanese language learning is known for being a big challenge, particularly for students that are not acquainted to its writing system. This language uses three different writing systems: kanji, kana (hiragana and katakana) and roma-ji. Kanji are Japanese ideograms brought and adapted from China. Kana are two syllabic alphabets – syllabaries – auxiliary to the language. Hiragana is used to write words whose kanji is rare or does not exist, e.g., greetings, adjectives, verb endings; and also as furigana (yomigana/rubi): a little subtitle inserted beside a kanji to reveal its reading. Katakana, in turn, is used to adapt foreign words and names to Japanese idiom, and also for emphasis of words. Roma-ji is the Roman alphabet, used in most languages in the world, including Portuguese, but in Japan it is barely used. Nowadays, the challenge of learning a language like this can be minimized by the use of different technological/computational artifacts that can be suitable for different kinds of students and their needs. A trend for language learning is the use of Mobile Assisted Language Learning –, a derivative subset from mobile learning and Computer Assisted Language Learning – which refers to an approach of learning that is somehow assisted by the use of a handheld mobile device (e.g., mobile phone, smartphone and tablet). In previous work, we identified the state of the art [35] of the subject-matter by the scientific literature, that presents computational solutions developed to support Japanese language learning. We also identified the state of the practice [43] focused on Computer Assisted Language Learning web tools for Japanese, representing the most used web tools, and evaluated their behaviour when trying to use them in an Android environment, observing multimedia

39 Este capítulo é uma versão revisada do trabalho [62], já publicado e apresentado. 55 Chapter 4. State of the Practice Android

disturbance caused by pure migration, finding that it is not ideal just to run these kinds of tools on Android operating system, because of incompatibility issues. After the accomplishment of these researches, it was perceived a convergence of computational solutions – software – for mobile platforms, also corroborated by the previous works [35][43], whose finding added more motivation to the achievement of the present work. In this chapter, we present an evaluation, from the Human-Computer Interaction (HCI) perspective, of diversified applications (Apps) developed specifically for mobile devices that use Android operating system. This evaluation was conducted in order to identify the most common/used design and interaction requirements/features. As the main result, it indicated the possibility of developing an App that can cover the needs of most of Japanese language students with access to an Android mobile device, based on requirements extracted with the help of the Semiotic Ladder framework. This chapter is organized as follows: Section 4.2 summarizes Mobile Assisted Language Learning; Section 4.3 presents the theoretical and methodological bases that ground this work; Section 4.4 introduces the evaluated Apps based on the Semiotic Framework, classified and organized in groups; Section 4.5 discusses the results under the HCI viewpoint; and Section 4.6 presents the final remarks.

4.2 Mobile Assisted Language Learning

Mobile Assisted Language Learning (MALL) is a combined subset of mobile learning (m- Learning) and Computer Assisted Language Learning (CALL). Various CALL applications were presented in [43], a work focused on web CALL tools specifically for Japanese language learning and the possibility of using them on a mobile environment; and its results led our research to MALL and Android operating system. For introducing MALL, now we present some academic works that exemplify the use of this kind of resource. Edge et al. [40] developed a flashcard system focused on the possibilities mobile offers. Their application, called MemReflex, is an adaptive flashcard application that gives the user a fast feedback, by the means of quick succession of retesting new cards/items. MemReflex works with audio features as well as text, and their research demonstrates enhancement of user’s skills in the target language, even when they were not concentrated completely on the application. Another interesting research is Tian et al. [50], that presents two culturally inspired learning games to help Chinese children to learn Chinese characters. The games are called Multimedia Word and Drumming Strokes; in the former, children must recognize and write correctly a certain Chinese character based on hints provided for pronunciation, a sketch, a photo or other multimedia context; and in the latter, children must write a character correctly with the correct stroke order, practicing the Chinese characters writing ability. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 56

Wong et al. [52] make a study over MALL that highlights “creative learner outputs” and contextualized meaning making, that is also seamless learning-inclined. They emphasize the importance of personal and social features mobile provides, and the good influence it can be in language learning. Their approach only evaluates in-class use of MALL. Zervas and Sampson [53], propose an IEEE LOM Application Profile – describes educational resources with metadata – intended to be used for tagging educational resources that are suitable for language learning and also supported by mobile devices. Magal-Royo et al. [42], demonstrates the availability of using multimodal environments, by both students and teachers, for accomplishment of testing and examination activities, in the field of language learning, showing that through mobile phones the levels of usability and accessibility can be improved, impacting on the development of new language learning methods. Still, Verma [51] describes Mobile Assisted Word Learning (MAWL), which is an augmented reality based collaborative interface that provides the user to learn new words by means of a smartphone, an approach based on CALL and MALL. MALL applications have advantages, such as the possibility of studying at almost any time and place [53] since the student is holding a mobile device; it creates great support for helping fixing content previously learnt, and even to actually learn new content, all in a customized way for each student. With the rise of Android operating system and the use of Android devices, everyday new Apps are available on Google Play® (olden Android Market®). MALL Apps for Android related to Japanese are varied, and generally purpose specific, even several of them with a couple of simple functions. They can be dictionaries, games, tutor systems, exercises with feedback, and are destined for different publics: children, adults, foreign and even Japanese natives. After seeing the advantages for the use of MALL applications, corroborated by the authors cited above, in the next sections we present our methodological and theoretical basis for this work, and then the results of the research about Android MALL Apps for Japanese language.

4.3 Methodological and Theoretical Basis

The main steps for conducting the research were: i) identification of different Android MALL Apps for Japanese language; ii) 30 minutes of exploratory use for each App; iii) classification and grouping similar Apps; and iv) evaluation of design and interaction features. The evaluation of the Apps will be made all manually, because this way it is possible to detect details from a user perspective, what would not be possible automatically. All the selected Apps are partly or totally free, and for the former, it was evaluated the free content. The specifications of the device used in the study are: Tablet Genesis GT-7200 with Android 4.0.3 – Ice Cream Sandwich, CortexA8 1.2GHz GPU 400MHz, screen size 7’’ (800x480). 57 Chapter 4. State of the Practice Android

It is important to highlight that in this work we do not present an exhaustive list of Apps. Once the number of application is expressive, we selected the most used Apps, based on the number of downloads, positive rating and indication of connexion at the Google Play®, in order to analyze and identify the most common patterns and requirements necessary for the development of a new App that would cover all the important aspects identified. Along the chapter, it is described whether the explored tool is available in Brazilian Portuguese language (PT-BR), since this identification is relevant to our research project context. In order to ground our research, we draw on the Organizational Semiotics (OS) theory. OS is a theory devoted to the study of organization using the concepts and methods of Semiotics, that comprises every organization as a system of signs. The OS offers methods and artifacts that allow the understanding, modeling and constructing of computing technologies. In this work, we used the Semiotic Framework (Fig. 5), an artifact from OS, for the requirements analysis of the Apps. Also known as Semiotic Ladder, this framework is an artifact composed by six levels that offer different points of view of various aspects of a given domain. The six levels of the Semiotic Ladder are briefly described as follows [40]:  Social World: in this level, it is observed the relationship between the use of signs and the effects it can cause on the social context;  Pragmatics: the pragmatics level focuses on the purposes of the signs. It evaluates the relationships between signs and the behavior of agents;  Semantics: the semantics level considers the relationship between a sign and what it refers to;  Syntactics: in this level, we can analyze complex structures of a language; it is related to rules of composing these complex signs from simple ones;  Empirics: the empirics level embraces statistical properties of signs, studying features like coding, optimal signal transmission and entropy measurement;  Physical World: in this level, a sign is observed in its physical form, being evaluated features as shape, size and hardware, for example.

Fig. 5 - The semiotic framework (c.f. [40]). Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 58

The methodological and theoretical basis in OS has been used on HCI researches in various works. For instance in the context of: social software design [47]; social networks [46][49]; applications [48] and physical devices [44][45] for the interactive digital television; systems for supporting problem solving and decision making [39], and so on.

4.4 Results

In this section, we describe the identified and evaluated Apps organized in the following sequence in the text: Tutor System Apps, Game Apps, Quiz Apps, Dictionary Apps, and Flashcard Apps. It is important to highlight that a flashcard or flash card is a set of cards bearing information, as words or numbers, on either or both sides, widely used as a learning drill to aid memorization by way of Spaced Repetition System (SRS40). There are various SRS computing systems, and for Japanese language learning, the most broadcast one is Anki [38]. Also, it is pointed out the main design and interaction features that are common to almost all the Apps, leading to a set of interface/interaction requirements for the development of a new Android MALL App. In order to facilitate possible searches at Google Play Store® for Apps or even to find the developer himself, we describe next to each App, the name of its developer. Also, it is important to report that some pieces of text in Apps’ description are highlighted in cyan, being them the identified requirements that will be explored later. Tutor System Apps are the most complete Apps in this classification, and they may embrace other kinds of Apps as aggregated features to provide a large learning environment for the user. Tutor Systems, in general, present content in an educational way, offering to users new content for them to learn. It uses features like tabs to present the new content, several of them with audio support, tabs for the students to review the content, and some of them present exercises for them to practice, with feedback, corrections and incentive to try again. Some tutor systems are as follows. Hello Hello Japanese (by Hello-Hello) is a tutor system with text and audio lessons, flashcards and notes. It is linked to the course by Hello-Hello company of language learning that allows users to interact with native speakers of their target learning language, also offering other language lessons than Japanese. This App offers only the first lesson of each language free, and the feedback is given by users when they submit the responses for their exercises. It is possible to use this App in PT-BR language. Similarly, there is also Busuu (by Busuu Online S.L.), which is a tutor system that offers several levels of courses with text, picture and audio lessons, revisions

40 The SRS is a system to study a subject systematically, using flashcards, repeating/showing more times the topics – cards – that are more difficult for the student to learn, and programming the most easy to appear more spaced in time. 59 Chapter 4. State of the Practice Android

and exercises with feedback. It is attached to the course by Busuu community, which is an online free community for learning various languages with interaction to native speakers. Japanese Hiragana Writing (by Teachersparadise.com) is a tutor system to teach hiragana alphabet. It can show a skeleton for the user to fill drawing, with the options of subtitle, arrows showing how to draw, or just the picture with no clue, all of them with audio available. This App does not present feedback to the user. On the other hand, Learn Japanese Numbers (by Ajeic) is a tutor system to teach Japanese numbers from 0 to 9999. This App offers basic lessons, and also, tests as flashcards in three levels (first level is from 0 to 99, second from 0 to 999, and third from 0 to 9999), in addition to a listening practice, both showing the answers as means of feedback. The tutor system Obenkyo (by Atomusk) teaches kana, numbers, kanji, vocabulary, particles and grammar. It also contains exercises like multiple choice, drawing and recognizing words and characters, and it is also available in PT-BR. The App Japanese Conversation Lite (by Saora Inc.) is a tutor system that teaches Japanese everyday phrases with text, English translation, audio examples and voice recognizing to evaluate user speak (few examples in the free version, and only available with Internet connection). This App counts with a large amount of audio content available free. JLPT Practice (by Saora Inc.) is a tutor system with focus in studies for Japanese Language Proficiency Test (JLPT41) exam. It contains a kanji dictionary, flashcards and also a sample test with feedback (its free version is a demo of a JLPT Level 2 test). There is also JA Sensei (by Raphaelw), which is a very complete tutor system with lessons about kana, kanji, vocabulary, verbs, particles, among others, and offers exercises with audio and drawing recognizer features, having a large amount of free content. Learn Japanese Language (by KHJLAB) is a tutor system with a childish interface, that teaches kana alphabets – with audio and mnemonic pictures features – and basic sentences using just kana. On the other hand, Japanese Verbs (by Szymon Trapp) is a tutor system focused only on Japanese verbs. This App teaches about all kinds of verbs, and also offers a verb dictionary plus a customizable quiz about the learnt content. Still, TenguGo Kana (by Tengulogi) is a tutor system with text and audio lessons and quiz about kana. Kana Draw and Kanji Draw (both by Lusil) are two other examples of tutor systems. The former is a tutor system that teaches how to draw kana correctly by offering the picture for the user to fill in, and providing immediate feedback about the kana strokes; Kanji Draw is the kanji version for Kana Draw. Some more examples are Learn Japanese Kana Dr Moku and Learn Japanese Essentials Free (both by Bob Byrne). The former is a tutor system with a childish interface, and presents kanas with English mnemonics associated to each one of them. It also offers quiz with immediate feedback. The other App by Bob Byrne, Learn Japanese Essentials Free, it is also a tutor system, but this one is focused on day-to-day phrases and just offers text lessons – phrases – with audio and English/Spanish translation.

41 The JLPT is an exam that evaluates and certifies Japanese language proficiency for non-native speakers. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 60

GettingThere (by SJC Linguistics) is a tutor system that teaches basic conversation phrases with text – kanji or only kana – and audio. It also offers exercises with the answers, but no kind of feedback. There is also Learn Japanese Deluxe (by MobileFusion), a tutor system with various modules, some of them in the free version, which includes vocabulary trainer, memory match from kana to English and some flashcards. One more example is Kana Mind (by Zeh Fernando), which is a mix of kana tutor and game, where the player is given a table with the roma-ji character and six tables with kana options. The goal is to pick up the corresponding kana. We also have Japanese Numbers Free (by PayToPun), focused on teaching Japanese numbers. It provides explanation about number’s kanjis and combinations, as a number converter from roma-ji to kanji, and two kinds of quiz for training. Other examples of tutor systems are: EasyKanji (by Micarisoft), Learn Japanese Quick (by Toy Box) and Learn Japanese (by Bravolol). EasyKanji is a tutor system that teaches from kana to kanji, showing readings, examples, offering spaces for annotations and kanji writing training. EasyKanji also offers a customizable quiz. This App has a stylish interface and allows the user to custom some view elements. Learn Japanese Quick offers lessons about various kinds of subjects in Japanese, but only three available on the free version. Interesting, this App offers a Study List and, then, several sorts of exercises to test users’ learning, including listening, reading and speaking skill tests, as well as vocabulary. It uses text, picture and sound features. Last, but not least, there is Learn Japanese that provides several example phrases with text (kana and kanji), audio and English translation. It offers the options of copying text in Japanese, English and roma-ji, and the user can also share content on Facebook®. Game Apps is the kind of App intended to unite learning and fun, in the most of times trying to mask the educational features to keep user motivated. They try to provide learning with fun, generally offering scores, records storage, and levels for the user to get motivated to play and unlock new content or more difficult ways of playing. In general, games do not offer explanation of content, different from Tutor Systems, but some of them can offer the player some clues of how to achieve the goal. Following, we present some identified and evaluated game Apps. It is important to highlight that these Apps are not just games in the Japanese language, but games that are focused on helping on the learning of the Japanese language. NihongoUp (by Philip Seyfi) is a game in a form of a multiple choice quiz where the answers come in balloons and the player has to blow up the one with the correct answer. There are modules of practicing/playing about kana, kanji, vocabulary and grammar, each one of these allowing the player to choose between classes of words; some of them are classified according to JLPT levels. The game offers immediate feedback during the game and final statistics. Another example, JaLaGa (by Joshua Jamison) is a very simple game App where the kana is shown and the user has to choose the correct roma-ji by multiple choice. At each level, more characters are added. There are no customizable options. 61 Chapter 4. State of the Practice Android

In the game App Kanji Writing Game (by Abraham Macias Paredes), the player is given English meaning, on and kun readings, and the goal is to draw the correct kanji. It is also possible to use it as a kanji searcher, drawing. This App has functions available in PT-BR, but the translations to the Japanese words are just available in English. One more example is Hiragana Tap (by Boon-Hau Teh), a game with a childish interface, where are presented several tables containing hiragana characters, and the player shall tap them in the hiragana alphabetical order. The game saves the score with the player’s name. Resembling it, there is Katakana Quiz (by Maxime Cardinal), but with a more simple interface and katakana characters instead of hiragana ones. Hiragana Memory Game (by Eleven11) is a memory game where the player has to turn tables to find the hiragana and its corresponding pair, which is another table with its roma-ji. It does not present score or time, just exhibits the number of attempts at the end of each game. The App Aiue Onigiri (by Makorino) is a childish game only in hiragana, that is indeed for Japanese kids to form words, but it can be used as a vocabulary training for foreign students. We also have Hiragana/Katakana Drills (by Robert Muth), a game where the player is presented a kana and has to select the relative roma-ji in less than the given time. Whilst the levels increase, more kanas are added to be presented. The score is simple, based on number of hits, and feedback of errs is imediate. And the game App GamuProg hiragana (by GamuProg) is a game to practice hiragana skills. A card with a Hiragana character is shown and the player has to select, between eight other cards, the one with the roma-ji relative to the former. The score is pointed out at the end, detailing the statistics of the game. The App Colors in Japanese (by Pibeans) is a game App for the user to practice reading and listening of the names of the colors in Japanese. The Game offers a practice mode for the user to get used to the colors that will be presented. Colors in Japanese do not offer score or levels, it is a very simple game App. Another game App is Touch Japanese Hiragana (by Hidanosatoyama), where the player is presented a character, has to recognize it and touch the same character between several that are presented. This game can help the player to practice hiragana symbols recognition, improving the player’s reading skills. Quiz Apps use is very simple. They are more like a test for more expert students, providing a challenging ambient for users to prove their knowledge about the subject. This kind of App can present a Q&A scenario, where the answers can be multiple choices with text, picture, audio or, in this case, drawing skills of kana and kanji characters. The identified and evaluated quiz Apps are presented below. Kanji Recognizer (by Nikolay Elenkov) is a kanji writing quiz, where the user has to draw correctly the characters that are asked. It is possible to configure the range and grade of the kanji to be on the quiz. At the end of each draw, it is exposed whether the user got it right or wrong, and at the end of the quiz, statistics are shown, allowing users to see their mistakes. There is also a training mode where it is shown the lines and numbers indicating the stroke order, while in the Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 62

quiz mode, it is only given the meaning and readings of the asked character. Another example is JLPT Master (by JCT Co. Ltd.), an App for advanced Japanese students; it offers proficiency tests with questions of multiple choice completely in Japanese, and feedback at the end of it, being also possible to use the whole App in Japanese language. The App Conjugation Japanese (by Tengen.Mobile) is a quiz App that offers exercises about verbs conjugation in various forms – e.g., nai, ta, i, and te –, each test with 50 questions and immediate feedback. Still, Japanese Trainer (by Kozhevnikov Alexey) is a simple quiz App that allows users to practice their knowledge about kana with a very simple question and answer system with immediate feedback. This App offers no statistics or score. Elementary Kanji (by Casmo Games) is a quiz App for kanji practice, with the options from first to sixth grade kanjis (i.e. kanjis teached since first until sixth grades of Japanese schools), with immediate feedback and stats, and allows the user to see the readings and English meaning to each kanji. Also by Casmo Games, Master Katakana is a quiz where the user is presented a katakana character and has to enter the relative roma-ji. There is a temporary status of correctness that lasts until the end of each training session. Another examples of quiz Apps are Kana Tutorial (by Karanagai), Katakana Learn Experiment (by Legendarya/Imaginactiva) and Kana by Hand (by Olof Sjöbergh). Kana Tutorial is a quiz App only for training kana, presenting the kana for the user to choose the roma-ji or vice versa, with immediate feedback. Katakana Learn Experiment is a quiz App where the users can evaluate their abilities of reading, listening, and writing katakana characters, by three kinds of questions. And Kana by Hand is a quiz App where a letter in roma-ji or an English word is presented and the user has to write the corresponding hiragana drawing it on the screen. The feedback is immediate and the progress is saved, so the users can see what characters and words they already got right. The App Japanese Grammar Free (by Eltsoft) is a quiz App that offers six levels, since presenting all questions in roma-ji (Beginner), the other levels present questions all in Japanese, upgrading the levels of the kanjis whilst the levels increase. The questions are of multiple choice and the user can access each level one time for free; to keep accessing, premium has to be bought. Another quiz App is Japanese Kanji Quiz (by Chamicat), a quiz divided by Japanese grades, where each grade has levels inside. The player is presented a kanji and the user has to select the correct kana reading, between four options of choice. Dictionary Apps are the simplest ones. These Apps are digital dictionaries where the user can easily consult words or characters, since they provide a language database with various possibilities of search manners. Japanese dictionary Apps can offer search by roma-ji, kana, kanji, stroke count, kanji radical, and even give words in another language as entry or response. The dictionary Apps identified and evaluated are presented next. JED Japanese Dictionary (by Attila Korompai) is an App that allows the user to download English/French/Spanish to Japanese dictionaries and offers search – kanji by radical, words, and 63 Chapter 4. State of the Practice Android

expressions – in roma-ji with results in kanji, in addition to a hiragana and katakana table. There is also Kabuto (by Gakusoft), a kanji dictionary with the options of searching by words (English or Japanese), kanji text, kanji radical and number of strokes. Another is WWWJDIC for Android (by Nikolay Elenkov), which is an Android version of the web tool WWWJDIC, a Japanese dictionary; and this App depends on Internet connection to work. The App Makimono Japanese Dictionary (by Remo Arpagaus) is a very large dictionary App, with search options by kanji, roma-ji, kanji radical, English meaning, and that contains a lot of example phrases for the searched word, with translations for various languages, including English, Portuguese, Spanish, German, Russian and French. Kanjidict (by Spartan Entertainment) is a kanji dictionary App in which the user can select the kanjis to be shown by JLPT level and/or grade, but there is no search tool. It offers stroke order indication and other information about the kanji, with English meaning. This App has made an attempt to be available in PT-BR, but it was not concluded. Also by this developer, there is Japanese Dictionary, which offers search options, several information about a found kanji and automatic conversion from typed roma-ji to kana. The same issue with PT-BR language occurs. One more example of dictionary App is Jishop (by Seva Alekseyev), that provides a complete kanji dictionary for a period of 30 days of trial. After that, it is only possible to access the called “school kanjis” for free, i.e. the list of 1006 kanjis learnt in Japanese school grades. Some features are search by roma-ji, audio readings, kanji dismemberment into radicals, stroke order diagram. Jishop also offers a very simple study option with drills and flashcard, but its main function is dictionary, justifying the classification. Still, we have Japanese-English Dictionary (by App For All), which is an App that provides a dictionary with search by words, kanji, English meaning. It is based on the web tool Denshi Jisho and does not work without Internet connection. Flashcard Apps usually consist of the simple use of flashcards to present and review content. Basically, they present content in flashcards and save the user’s progress along the learning process. They often use text and audio features for the sake of fix content by the user, and can use various multimedia resources. Follows the identified and evaluated Apps of this kind. Color Kana (by Huangtiancheng) is a simple flashcard App where users study kana with text and audio, and mark each kana with a color depending on the facility they had to read/remember. Another example is Simple Japanese Gojuon (by Limitfan), a simple table App, with a sort of flashcard that can show pronounce (with audio), stroke order (animation) and example of use in a word, of kanas. It also contains a feature that gives a word to be spoken and recognizes user’s pronounce as an audio entry, but it depends on Internet connection. AnkiDroid (by Nicolas Raoul) is the Android version of Anki, a SRS tool very known by Japanese students. It allows the users to create their own deck of flashcards and has the possibility to sync with their desktop and web versions of Anki, for the decks to be updated. Anki offers plugins specially for the study of Japanese language and AnkiDroid is available in PT- Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 64

BR. Another App is Spare Time JP (by Aikenau), that teaches kana’s basic sounds showing the list in alphabetical order. When the user touches a character, the relative sound is played. The Flashcard App Japanese Vocabulary (by Bravolol) is an App that contains flashcards organized by subject. Each flashcard presents two sides, one with the Japanese word, and another with English translation, roma-ji and relative audio. The user can also mark flashcards to a favorite list. One more example is Qwiz Hiragana (by Ed Halley), a flashcard App that teaches Hiragana characters and saves user progress to set the cards that will appear in the study each time. This App also provides the user to earn tickets to redeem power-up prizes. FlashKana (by Plattom) is a very simple flashcard App, which only contains two-sided cards, one side containing a kana, and the other side, its roma-ji and alphabet indication (hiragana or katakana). There is no use of SRS system, no saving of progress and no feedback. It is possible to choose what cards to see, and whether they will appear in alphabetical order or random. Similarly, there is Kana Cards (by GamEnlight Studios), but with the difference that the latter offers kanji cards too. Another one, the flashcard App Flashcards for Kids (by MOZ Android), brings animal pictures with audio and text (kanji and kana only), but also does not use SRS or progress saving. The App Japanese Flashcards (by uTech Studio) provides a SRS system and cards with kana/kanji in the front side and roma-ji/English meaning on the back side; it offers four decks for training since kana to words and kanji. The flashcard App Kanji Flashcards (by Joseph Daverin) provides kana and kanji flashcards for training, with the option to select the alphabets, and also the kanji by grade. There is a practice mode, where the first side of the card shows all the information, and also a quiz mode, where the user sees the kana/kanji and touches the screen to see the response and mark the user hit or fault in knowing the character’s information. Still, JFCards – Kanji Flashcards (by Plamen Parvanov) provides cards, one side with kanji, the other with English translation. For each card, there is also an option to see more information about the character. When conducting the search for MALL Apps for Japanese, several of them were found. The authors are aware that this list can be very expressive, but as was explained before, the Apps were selected under criteria presented above (e.g., number of downloads, positive rating and indication of connexion at the Google Play®). Still, some of the classifications, like Game and Quiz Apps, present just a few relevant tools that we intended to introduce in our research; searches for these kinds of Apps at Google Play® would not exhibit much more expressive results. It is also important to emphasize that this classification was made after manual analysis of each App, under the methodology presented in Section III; because of that, there is the possibility that some table classification may have been affected by means of features that remained “hidden” until the end of exploratory time. Table 4 summarizes 64 Android MALL Apps that were identified and evaluated. This table presents the column names: “Free” that indicates whether the App’s content is available (T)otally 65 Chapter 4. State of the Practice Android

free, or (P)artly (i.e. some features are only available if the user gets the full version of it); “Lang. skills” that indicates language skills embraced by each App, i.e. (R)eading, (W)riting, (S)peaking, and/or (L)istening; “BR” that indicates whether the App is available in PT-BR, i.e. (Y)es, (N)o, or (P)artly; and “Audio”, where “I” indicates use of input audio, “O” indicates output, “I & O” indicates use of both and “-”, that represents no use of audio features for learning purposes. Table 4 - Android MALL Apps Identified and Evaluated Type/Icon Name Ver. Free Lang. skills BR Audio Hello Hello Japanese 1.1 P R, S & L Y O

Busuu 2.5 P R, S & L Y O

Japanese Hiragana Writing 6 T W N O

Learn Japanese Numbers 5.50 T R & S N O

Obenkyo 3.3.1 T R & W Y -

Japanese Conversation Lite Natsu-12b P R, S & L N I & O

JLPT Practice 1.4.2 P R N -

JA Sensei 2.6.1 T R, W & L N O

Learn Japanese Language 1.0 T R & L N O

Japanese Verbs 2.1.0 T R N -

TenguGo Kana 2.0.10.128 T R & L N O

Kana Draw 2.3 T W N -

Tutor System Tutor Kanji Draw 3.3.1 T W N -

Learn Japanese Kana Dr Moku 1.0 P R N O

Learn Japanese Essentials Free 1.0 P R & L N O

GettingThere 2.2 T R & L N O

Learn Japanese Deluxe 1.7 P R & L N -

Kana Mind 1.5.71 T R N -

Japanese Numbers Free 1.09 T R N -

EasyKanji 1.1.0 T R & W N -

Learn Japanese Quick 1.3 P R, S & L N I & O

Learn Japanese 2.2 T R & L N O

Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 66

NihongoUp 1.0.1 T R N -

JaLaGa 1.2.0 T R N -

Kanji Writing Game 2.0 T W P -

Hiragana Tap 1.2 T R N -

Katakana Quiz 1.00 T R N -

Hiragana Memory Game 1.6 T R N -

Game Aiue Onigiri 1.0.5 T R & L N O

Hiragana/Katakana Drills 13.demo T R N -

GamuProg Hiragana 2.3.3 T R N -

Colors in Japanese 2.0 T R & L N O

Touch Japanese Hiragana 5.0 T R N -

Kanji Recognizer 2.3 P R & W N -

JLPT Master 1.09 P R N -

Conjugation Japanese 1.0 T R N -

Japanese Trainer 2.1.0 T R N -

Elementary Kanji 2r T R N -

Master Katakana 1r T R N - Quiz Kana Tutorial 1.1 T R N -

Katakana Learn Experiment 1.00 T R, W & L N O

Kana by Hand 1.3.1 T W N -

Japanese Grammar Free 1.0 P R N -

Japanese Kanji Quiz 1.0.9.4 T R N -

JED Japanese Dictionary 0.5.5 T R N -

Kabuto 1.04 T R N -

WWWJDIC for Android 2.2.5 T R N - Dictionary Makimono Japanese 0.3 T R P - Dictionary

67 Chapter 4. State of the Practice Android

Kanjidict 1.4b T R P -

Japanese Dictionary 1.3 P R P -

Jishop 2.10.0 T R & L N O

Japanese-English Dictionary 1.1 T R N -

Color Kana 2.2.1 T R & L N O

Simple Japanese Gojuon 2.5 T R, W, S & L N O

AnkiDroid 2.0 T R & L N O

Spare Time JP 1.0 T R & L N O

Japanese Vocabulary 2.2 T R & L N O

Qwiz Hiragana 1.3 T R N -

FlashKana 1.0.1 T R N -

Flashcard Kana Cards 1.6.3 P R N -

Flashcards for Kids 1.0 P R & L N O

Japanese Flashcards 1.02 T R N -

Kanji Flashcards 1.5.5 T R N -

JFCards – Kanji Flashcards 1.3 T R N -

After identifying and evaluating the Android MALL Apps, we accomplished an analysis in the domain of MALL Apps for Japanese based on the Semiotic Framework. Table 5 presents some questions involved on the analysis according to the six levels of the Semiotic Framework. Table 5 - Semiotic Framework in the Analysis Domain of MALL Apps for Japanese. Level Questions Involved

What is the profile of the audience that makes use of MALL Apps for

learning Japanese? Is there a social separation because of the need of owning

or having access to an Android mobile device? Is there any kind of Social Social World interaction with other users (e.g., social features)? Can the Apps bring some idea of contempt about the role of a traditional teacher? Is it effectual/interesting to integrate the use of MALL Apps to traditional

learning? Is this integration supported by teachers? Are the Apps used as a

ics complementary manner? Do students resort to other resources (e.g., books Pragmat and websites)?

Are the MALL Apps’ interface and interaction features making sense for the students in a learning perspective? Does the user regard the MALL Apps as

a relevant way of language learning?

Human Information Functions Information Human Semantics Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 68

What are the languages used to develop Android Apps? What are the technologies involved? What are the tied guidelines to follow (if there is

any) related to MALL Apps?

Syntatics

Do the Apps need to have Internet connection at the moment of its use for it

to function part/completely? What is the required bandwidth? Empirics

The IT Platform IT The Are the Apps making use of hardware features such as the accelerometer? What are hardware requirements (e.g., performance, energy consumption, and devices) relevant to the application? Are all the physic environments

World appropriate for the use of some features of MALL Apps, e.g. voice Physical Physical recognizer?

Through the Semiotic Framework it is possible to analyze different aspects of the same problem, favoring the identification of questions that might be obvious, but that were only clarified when the use of the artifact was performed. Thinking about the Physical World, it was possible to perceive a lack of use of a very known characteristic of tablets and smartphones – the accelerometer – by Japanese MALL Android Apps as a way of interaction. The Empirics brought to scene a frailty present in some Apps, which is the need of Internet connection to run some functions or even the whole App. Syntatics raised questions directly related to the programming step of an App development, and with these three levels that form the IT Platform, directions in this meaning are found for future development. Analyzing the Human Information functions, we have Semantics, which brings reflection about the purpose and validity of using Japanese MALL Apps. Pragmatics, that questions about negotiating between teachers and students for using better the Apps; and Social World, which considers the social questions for applying the use of Apps for learning (e.g., accessibility of devices). Follows, results from observations about interface and interaction features in the evaluated Apps, highlighting what is most common based on the created classification in order to point out the main requirements that each one has. In this case, we are not focused on Android User Interface Guidelines [37] themselves, but in what is more common to the Japanese MALL Apps in specific. The Japanese MALL Apps, in general, may be very specific or very embracing in its content. They can focus on reading skills, presenting: i) texts for recognition (Fig. 6), and ii) texts in Japanese with corresponding roma-ji and translation (Fig. 7). Writing skills: i) recognizing user writing by the touch screen (Fig. 8); ii) offering tutoring of how to write a character (Fig. 9); and speak by capturing user audio and analyzing it (Fig. 7). Listening: i) playing a corresponding audio for a piece of text for the user to listen and learn the pronounce (Fig. 7), and ii) playing an audio and asking user to identify what was said in it (Fig. 10). 69 Chapter 4. State of the Practice Android

Fig. 6 - Reading skill quiz (Elementary Kanji App).

Fig. 7 - Japanese text with roma-ji and translation, audio teaching, and user speaking recognition (Japanese Conversation Lite App).

Fig. 8 - Kanji drawing recognition (Kanji Recognizer App).

Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 70

Fig. 9 - Animation showing the stoke order for the hiragana “ga” (Simple Japanese Gojuon App).

Fig. 10 - Listening skill quiz (Learn Japanese Quick App). Analyzing the aspect of interface, most of the Apps do not take into account contextual appearance – e.g., Japanese day-to-day relative pictures, and cultural features – which can take the user to a more immersive environment and improve language learning as well as Japan cultural knowledge. In the interaction features – as transition of screens and access to options –, organization of screens and buttons, most of the Apps follow Android UI guidelines, which allow different kinds of interaction but with good usability features. To perceive all these details, it was fundamental to evaluate manually every single App – following the steps presented on Section III –, whereas automatic evaluation does not allow us to assess the Apps with the perspective of a user. Hence, it was possible to notice their positive and negative points, and then perceive what features can be the most effective in the learning context and pleasant in a usability perspective, and how to work them together in the development of a new App. Table 6 summarizes the identified requirements from the description of the Apps (highlighted in cyan in beginning of this Section), as from the analysis of interface features by the Figures 6, 7, 8, 9 and 10. 71 Chapter 4. State of the Practice Android

Table 6 - Identified Requirements According to Apps’ Classification App type Identified features Audio corresponding to a piece of text Audio lessons Combining text, picture and sound features Custom view elements Customizable options Dictionary available on the App environment Drawing recognition

English/Spanish translation Exercises with the answers Explanation of content/subjects of the Japanese language Flashcards section Immediate feedback Tutor System Tutor Play audio for the user to identify what was said Several sorts of exercises Showing readings and examples Skeleton of kana/kanji to fill drawing Text in Japanese with corresponding roma-ji and translation Text lessons Text with example phrases presented in kana/kanji Voice recognition Customizable options

Immediate feedback

Levels increase

Score Game Statistics Time Indication of hits and errs Possible to configure Questions of multiple choice Save progress Score Statistics

Quiz Status of correctness Texts for recognition Training mode Upgrade of kanji difficulty whilst the levels increase Use the whole App in Japanese Writing recognition Information about the kanji Kanji dismemberment into radicals

Search by: words (English or Japanese), kana/kanji text, kanji radical, and number of strokes Search tool Stroke order indication/diagram Dictionary Translations for various languages Work offline Work online bringing more features Animated stroke order Feedback Option for cards to appear in order or randomically Possibility to sync cards with tools of other platforms Practice mode Save progress

Flashcard Select kanji by grade Tutor of how to write a character (kana/kanji) Use of SRS Voice recognition

Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 72

Table 6 puts together the most common features/requirements present in the identified Apps, classified according to the Apps’ type. It is interesting to notice that some requirements are common to more than one kind of App. For example, “Feedback” present on Tutor Systems, Games and Flashcards, and also in Quiz Apps in the form of “indication of hits and errs”. Quiz and Flashcard Apps share the “training/practice mode” and “saving progress” features, while Quiz and Games ask for a “statistics” requirement. Another observation is about Tutor Systems, since some of its features are actually other kinds of Apps, i.e., a “dictionary” option, a “flashcard” section, quizzes for “exercises”. Game Apps presents some specific requirements, despite they are not manifold, as “score”, “time” and “levels increase”, which are characteristics usually assigned to games in general. Another feature that were only presented in one kind of App is “kanji dismemberment into radicals” that is present on dictionary tools; this is a very theoretical feature, what can explain its rareness in more practical kinds of Apps.

4.5 Discussion

Here, we present some challenges and directions for the development of new Japanese MALL Android Apps. It is significant to state that the use of the OS Framework brought several questions that were fundamental to reach the requirements elicitation, since it divides the domain in various levels, facilitating analysis by pointing out direct questions for each level. First of all, it is important to notice that these Apps can be very purpose-specific as well as very embracing. For the former, it is common to find a few functions and requirements, while for the latter, more often we find lots of resources with multimedia and, in some way, a condensation of various kinds of these Apps, i.e. gathering different features from the several kinds of Apps in order to generate a powerful tool with many resources. That is what we see in most Tutor Systems, that can include inside them some sort of quizzes, dictionaries, flashcard features, all of this mixed with text, picture and audio resources to form a big learning system. The other four classifications made here present more purpose-specific Apps, which can be very simple and still have their educational and fun values. These kinds of Apps are very good for practicing and reviewing content in a quick and simple way, or in the case of dictionaries, to quickly search for content that the user has doubts about. It is even more practical in the cases where there is no need for Internet connection. Despite the classifications and tied requirements may help the developer, giving him directions to follow, these paradigms can be against design and interaction innovations. Hence, it is important to follow some guidelines, but it is not mandatory every App, in each classification, to use the same flow, interface or other interaction designs pattern. It is important that the developer feel free to insert different features and put them under test, for the Apps to evolve. 73 Chapter 4. State of the Practice Android

Moreover, the language skills embraced by each App can be independent of its classification, but there are some patterns that can be noticed, as reading is the most contemplated skill, being it prevalent alone in Game, Quiz and Dictionary Apps, and massively present on the other two kinds (Tutor System and Flashcard Apps), accompanied by other skills. Tutor Systems tend to embrace more skills, since they are generally more complete Apps with lots of different features. Speaking is the less covered skill, and it may be assigned for the sake that it requires more complex programming to evaluate user inputs, since they can be very different and still be right. Another issue about practicing the speaking skill is the fact that it naturally imposes usage restrictions, e.g., public places are not favorable, because of ambient noise, the need of being in silence or speaking softly. Listening and writing are, respectively, the second and third more reached skills. This can be explained by the fact that, while a listening practice can be given by an audio play for the student to give the right answer in text or multiple choice (of text or image), a writing practice requires a more complex data processing for recognizing draw content, specially kana and kanji, since they have many particularities (as stroke order, stroke count, size and distance of strokes). In other words, reading and listening are the most contemplated skills in the MALL Apps because it is easier to deal with them from a computing perspective. The availability of Apps made to the Brazilian public is scarce, and because in Brazil the Japanese language is one of the most searched languages for studying, it represents an open market for the development of these kinds of tools. This work presents new results for the Latin America community, showing that the Apps focused on its public is still scarce; this represents opportunities for development of new Apps for supporting Japanese language teaching focused on this community. In the HCI field, it is important to highlight that the results achieved by this research demonstrate that the Organizational Semiotics can cooperate with Japanese language learning domain too. This work, therefore, complements the literature in the HCI field, showing the value of OS over the application domain researched in this work, i.e., interface/interaction in Japanese MALL Apps; a domain still not investigated under this perspective, and which demands informed and theoretical grounded discussion.

4.6 Conclusion

In this work, we identified various Japanese MALL Android Apps and created a classification for them according to their features. We also evaluated interface and interaction factors, based on Semiotic Framework – an artifact of Organizational Semiotic – in order to identify requirements for the development of a new tool. The interface and interaction features were evaluated starting from our classification of the Japanese MALL Android Apps, giving directions of how to Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 74

develop a new App based on the most common requirements used in each kind. It was possible to see the usefulness of the OS on the field of HCI directed for the domain of Japanese MALL Apps. As future work, we point out the development of a new App that contains suitable features, a development guided by its kind, depending on the intended classification, i.e., taking into account the classifications made here and the features and requirements detected can be helpful for the development of new Apps.

4.6.1 Decision Making

After this considerable survey, we realized that there is an aspect in Japanese language learning that, in general, is not contemplated by most of the available tools; it is the cultural aspect. So, we decided to give this aspect a special attention, and now we intend to research about how to insert this aspect in a CALL tool at this context. The next step of our research, instead of beginning the project of a new tool from now, is to study the importance of the cultural aspects on the learning of this language, for then create a new concept of a tool.

Chapter 5 5 Development of Karuchā Ships Invaders42

5.1 Introduction

The use of technological resources is present in most of the daily activities, and it is not different in the educational domain. These resources are being largely used informally for helping language learning in several contexts, this is called Technology Enhanced Language Learning (TELL), and it encompasses the use of computational tools for the same purpose, in literature called Computer Assisted Language Learning (CALL). Furthermore, there is the rising of educational games bringing a new way to study manifold themes, the so called gamification [67]. Gamification is basically to use game design elements in non-gaming contexts [54], to bring out motivational factors that are usually applied on games, to another activities, with the goal of arousing people to practice what would be boring activities, just for the enjoyment of doing that. Once these motivating factors are put in the context, and the tasks can become rewarding, it can cause positive effect on the user. An interesting approach about educational games is edutainment – i.e. education + entertainment –, which is to mask the educational piece of the tool with fun, sometimes with gamified learning activities, to make the learning to be a consequence of playing a game. Language learning is also included in the field of edutainment and gamification, in the attempt of making the process of learning a new language to be a pleasant activity for the student. In this context, for example, Pereira et al. [64] present a game for helping children learn Portuguese language. Japan is an island nation located in the Asian continent, which is, in culture ways, very different from occidental countries, as Brazil. Its language is the Japanese, a language that owns a

42 Este capítulo é uma versão revisada do trabalho [80], já publicado e apresentado. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 76

lot of particularities: it has three main alphabets, being two of them syllabics, i.e. hiragana and katakana (which together can be called kana), and the kanji, which are logographic characters adopted from China, and they also use roma-ji – Latin script – in some cases, for example, imported acronyms. This characteristic itself already defines a challenge and/or a barrier for occidental people that aim to learn about Japanese language. From the three main alphabets, the most important and basic to know is hiragana. Hiragana characters are the first taught in school for little children; furthermore, it is the most used alphabet: for words that do not have kanji, for adjective and verb endings, and also used for identify readings of unknown kanji, being inserted little hiragana letters by its side (in this case, called furigana). In previous chapters, we presented the state of the art of Japanese CALL tools which have been developed in the academic environment [35], the state of the practice of Japanese CALL web tools [43], and the state of the practice of Japanese Android MALL apps [62]. All these research brought questions that needed to be answered about Japanese CALL, and an important axis that remained unexplored is Japanese culture – and what is the importance of that influence on the design/development of a CALL tool – in the sense of providing improvement on the learning experience by giving the user a more immersive environment related to the target language. It was also perceived with these research that, in general, games for helping Japanese language learning are scarce, specially for native Portuguese speakers; and still, this kind of game usually do not take into consideration cultural aspects of the involved countries. This chapter introduces a CALL game directed to Japanese language learning, considering some aspects of the Japanese culture. In this step of our research, we intend to available a tool – a game – that can help Brazilians to get conversant to basic Japanese language elements, through an entertaining gamified way. The game was designed and developed by/for Brazilians, being necessary a study of Japanese cultural features, which is performed with the purpose of drafting elements to be used in the interface as well as in the different levels of the game, for contextual and educational purposes. The game developed here intends to teach and fix the main basis of Japanese language, which is hiragana alphabet, in addition to teach cultural words that can be written with these characters for learning vocabulary too. This chapter is organized as follows: Subsection 5.2 presents works and CALL games for Japanese language related to ours; Subsection 5.3 presents an overview of the Japanese culture and describes the influence of Japanese culture on the design of the game developed; Subsection 5.4 presents the developed game; Subsection 5.5 presents a discussion; and Subsection 5.6 concludes the chapter. 77 Chapter 5. Development of Karuchā Ships Invaders

5.2 Related Works and Games

Focusing on Japanese language learning, in the academic field, Stubbs [66] presents a Kana no Senshi, a desktop game that intends to teach Japanese beginner students to read kana quickly, for them to be able to start kanji classes efficiently. The game can analyze player’s performance, i.e. speed and accuracy, and with the progress, the enemies – that come like zombies – become faster, stimulating the player to read and identify kana faster too; this makes the game effective for learning and also fun. The author, based on the validation process of their system, defend that it has good acceptance by the users because it unites fun with learning. In the practical field, the Kana Invaders [61], the player has a ship too, but this one is controlled by the keyboard arrows; the enemies fall down similarly to [60]. When the player is sailing through the air and the ship’s sight meets the target, a signal is showed on that character, asking the player to type the corresponding roma-ji for it. One more desktop-based game is Kana Attacks [56], with the particularity that the player has to sail to meet the target, but when it is done, spacebar has to be pressed and a dialog box will appear showing momently the hiragana for the player to type roma-ji related. After typed the correct roma-ji, the screen with the ship and enemies comes back and a shot is fired to the typed letter. The web-based game Kana Invaders [60] allows the user to choose between hiragana and katakana at the beginning of the game; then the user is given a ship and enemies start to fall bringing characters with them. For the ship to shoot, the player just has to type the roma-ji corresponding to the enemy to be shot. Also, there is another web-based game with same name [59], which brings different approach, i.e. several enemies are displayed in rows and columns on the screen, each one with a hiragana character, and a roma-ji is given on the top left; the player just has to click pointing at the corresponding symbol, the click itself is the shot and the player has to hit the correct character; at the end of each level, time and accuracy are given. Another web-based game is Kana Drop [58], in which the player can also choose between hiragana and katakana characters at the beginning; the enemies – characters – come falling down and the player has to type the corresponding roma-ji and press the Enter key for the enemy to be destroyed. The game provides a view of what the player is typing, and also it is possible to erase the typed characters whether a mistake is perceived. The Kana Attack [55] brings the enemies as animals holding plaques with kanas written on it, and the player is given a hammer charged with a roma-ji that matches some of the enemies. The player has to hammer on the corresponding enemy related to the pronounciation charged on the hammer at the moment. The Kana Drop [57] brings a Japanese scenario where fixtures fall down bringing kanas, and the player has to type the corresponding roma-ji using the virtual keyboard of the device; for that, this game is more suitable for the iPad tablet. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 78

Table 7 shows a synthesis of the related works and games, with the name of the game, version, platform for which it is been developed, interface language, and information about if it is free. Fields with “-” means the information was not specified by the author/developer. Table 7 - Related Works and Games Ref. Name Version Platform Language Free [66] Kana no Senshi - Desktop English - [61] Kana Invaders 0.3beta4 Desktop (Windows) English Yes [56] Kana Attacks 0.91.0.13 Desktop (Windows) English Yes [60] Kana Invaders beta Web English Yes [59] Kana Invaders - Web English Yes [58] Kana Drop 1.0.3 Web English Yes [55] Kana Attack 1.1 Mobile (iOS) English No [57] Kana Drop 1.2 Mobile (iOS) English No

It is important to notice that all these works and games cited above do not contemplate cultural aspects of Japan, they only bring kanas and some of them bring interface elements that remind Japan. In this study, we intend to open space for cultural aspects attached to the most elements of the game that are possible. Next section we present some aspects of Japanese culture, opening possibilities of what can be used in this field to bring a more immersive and contextualized learning environment inside the developed game.

5.3 Overview of the Japanese Culture

We organized the Japanese culture in some main aspects, i.e. communication, clothing, feeding, arts, popular culture, beliefs, and people’s names, as described below. As Japanese culture is very large, we present an overall perspective of main points, not intending to embrace the whole theme, but to understand better the context and the usefulness of these elements on the game design. Related to its own language, communication between people in Japan has its particularities. Japanese verbs have several forms, representing levels of politeness, being the two basic forms, i.e. (i) dictionary form (informal), and (ii) “masu” form (polite). Depending on whom the speaker is talking to, one or the other form is more suitable, and the use of the correct form is deeply important, because it shows the respect due to the interlocutor. Still, in Japanese verbs there is no difference between present and future grammatical tense, as well as there is no gender or plural for substantives; besides, it is usual that some terms of a phrase are not said or written, to be tacit; all this making conversations plenty related to concerned context. The calligraphy – shodō (in roma-ji) – in Japan has also a tradition, and it is an artistic way of writing Japanese characters that contains several techniques, special brushes, inks and papers. Communication aspect is strongly present in the game, since it is its core. The main goal for what 79 Chapter 5. Development of Karuchā Ships Invaders

the game presented in this chapter was developed by our research group is to help the learning of Japanese language from the basis – hiragana alphabet –, that is, the beginning of communication. Japanese modern clothing is composed by western and Japanese clothing. Since the former is being largely used on the daily, the Japanese traditional clothing passed to be used in special events, like ceremonies, festivals and funerals. The Japanese traditional clothing contains several variations of a special garment that looks like a robe. They have kimono and yukata as the most known and used kinds, being the former a formal garment that differs from male to female, and for the women, it can also vary depending on age or marital status, for example. Yukata is a most casual version of the kimono, and can be used at home and inns, after bathing. There are more several garments in the traditional Japanese clothing, e.g. nagajuban, haori, haori himo, hakama, tabi, zori, and obi. Clothing aspects influence the pictures that appear on the game in every picture that contains a garment, and the words are present on the bosses. Japanese feeding has many traditional and known dishes. It is based on rice, soup, vegetables and meat (often fish), that can be uncooked, baked, fried and so on. Some known dishes are agemono, which is fried food; yakimono, grilled meat, poultry, fish and vegetables; nabemono, the baked ones; sashimi, the traditional raw meat, fish and seafood; and sushi, that is made of cooked vinegared rice combined with other ingredients, usually raw fish or other seafood. Japanese cuisine has also some known sauces, as shoyo (soy sauce), karashi (kind of mustard), mentsuyu, ponzu, warishita and wasabi, for example. Another particularity of Japanese feeding is its cutlery: instead of fork and knife, they use hashi, which are chopsticks used for eating; this is because they believe there is no good in having weapons on the eating table. The feeding aspect is also availed for bosses. The arts in Japan are sundry, including painting (kaiga – that embraces several genres and styles), sculpture (the traditional clay figures), ukiyo-e (woodblock print), ikebana (artistic flower arrangements), literature, being these all part of traditional Japanese culture. The modern popular culture contemplates Japanese movies (with famous genres as horror, tokusatsu, anime, and sci- fi); music – being the most famous J-rock and J-pop – that nowadays is spread all over the world through cartoons and games; and manga, the Japanese cartoon magazines. In Japan, most people do not follow a single religion; instead, they incorporate elements of several beliefs, and this is what they call shinbutsu shūgō. The main beliefs are Buddhism and Shintō, but there are a few representations of Christianity, Islam, Hinduism, Sikhism, Judaism, and Taoism, that are practiced with the religious freedom enjoyed by all Japanese. The arts, popular culture and beliefs are mainly present in the interface and interaction features of the game, as design and sound, and also in the cultural words (bosses). In Japan, people’s names is taken as a very important thing, there is a belief that a name’s meaning can influence personality and fate for the whole life of a person [65]. When a baby is born, the parents choose a name that can bring the desired fortune and/or behavior. Japanese also have their own numerology for people’s names, called seimeihandan, which can also Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 80

influence on the name’s choice. Japanese have also a preferred number of kanji in a name, which says that the sum of first name and surname should not have more than five kanji. Beyond the choice of pronounce, there is also the choice of the kanji that will compound the name, since several kanji can have the same reading. Futhermore, in Japan there is a list of allowed kanji to use in people’s names. This list is authorized by the Japanese Ministry of Justice and contains about 2,200 kanji. Another particularity is that, to have a surname, or a family name in Japan was, until Meiji era, a privilege of higher society classes. This cultural aspect in particular has influenced on the game story, for the choice of names with meaning and in the name of the game itself.

5.4 Karuchā Ships Invaders

The Karuchā Ships Invaders CALL game43 brings a story where a Brazilian city is being “invaded” by cultural Japanese ships, actually they are coming in peace and just want to be well received, and the player has to make the ships land safely in the Brazilian city. The name of the game was also inspired by a Japanese word, i.e. culture in Japanese is written in katakana as カルチャー and your roma-ji is karuchā. The game is invaders-like, and is also inspired in the work of Miranda et al. [63].

5.4.1 Plot, Gameplay and Features

Karuchā Ships Invaders starts when a couple of Brazilian exchange students go to Japan to learn about its language and culture. They have the mission of learn it and rehearse the knowledge to their friends who stayed in Brazil – which is the player –, because soon Japanese ships are coming to visit Brazil, and they will need Brazilian help to land; for that, it is necessary that Brazilian know about basic Japanese language and cultural aspects, so they can guide the ships to land safely in Brazil. Each day, the duo learn new things about Japan and communicate with the friends by means of text and pictures that are presented before some levels, teaching about new hiragana letters they learnt and new cultural experiences in Japan. The player has a laser to guide the Japanese ships to the ground, which works in a particular way, i.e. using Japanese language knowledge. Each ship is a visitor that comes with a hiragana character written on it, and the player has the power to help it land, by typing the corresponding roma-ji to load a shot of light on the laser. Whilst the laser is charged, it will direct the shot to the corresponding ship – i.e. a hiragana character or an icon that represents one aspect of Japanese culture – whose roma-ji was typed, and then the player just do the command

43 Available for download at http://code.google.com/p/karuchashipsinvaders/. 81 Chapter 5. Development of Karuchā Ships Invaders

for the laser to shoot the guiding light, which is to press the Enter key or backspace. If a ship is not well received (the player does not type its corresponding roma-ji in time), it hits the city and the player will lose a percentage from their hit points. Also, if the player gets to pass a level without allowing any ship to hit the city, the player wins 10% more of their hit points for the next levels until the end of the in progress game. Karuchā Ships Invaders is a game developed to help beginners in the Japanese language to learn the hiragana alphabet in a different and more fun way, and being more immersive in the Japanese culture and the differences between it and the Brazilian one. This game is intended to be played for firming content learned by other methods, as for learning new content by playing. Though, we provide a sort of tutor presenting a story with characters – the two exchange students – that introduce the player to cultural elements that will be present in the game (as bosses). It still can be played for those who are yet conversant to hiragana, for practice of content and evaluation of knowledge, and also for improving vocabulary, by means of the cultural words (bosses). The difficulty of the game contains three modes: easy, normal, and hard. They differ in the average time the ships and bosses fall down on the screen, being easy from 5 to 7 seconds between each ship/boss, normal is 4 to 6 seconds, and hard is 3 to 5 seconds. Withal, bosses’ falling speed is decreased depending on the number of characters they have, i.e., for example a word containing five hiragana characters will fall down slower than a word that contains three hiragana characters, according to the following formula, where “t” is the time of fall, “dm” is the difficulty mode, with the values 1 for hard, 2 for normal, and 3 for easy; and “noc” counts the number of hiragana letters used to form the word.

t  dm *3 dm  noc

The game, available initially in Brazilian-Portuguese (PT-BR) and English languages, is composed by 30 levels, each one adding a new “gyo” – i.e. column with a family of five or less hiragana characters in the Japanese alphabetical order – for the player to practice hiragana. Table 8 shows the 107 hiragana characters, each level of the game with the respective gyos (families/columns of each hiragana character, in alphabetical order), the new characters that are presented in each level and total number of new characters included in the level. Levels 15 and 30 are totally dedicated to cultural aspects, and that is why they do not contain new single hiragana characters. Table 8 - Basic Elements of the Game Hiragana (with its corresponding roma-ji) Level Gyo Character # 1 あ(a) あ(a) + い(i) + う(u) + え(e) + お(o) 5 2 か(ka) か(ka) + き(ki) + く(ku) + け(ke) + こ(ko) 5 3 さ(sa) さ(sa) + し(shi) + す(su) + せ(se) + そ(so) 5 4 た(ta) た(ta) + ち(chi) + つ(tsu) + て(te) + と(to) 5 Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 82

5 な(na) な(na) + に(ni) + ぬ(nu) + ね(ne) + の(no) 5 6 は(ha) は(ha) + ひ(hi) + ふ(fu) + へ(he) + ほ(ho) 5 7 ま(ma) ま(ma) + み(mi) + む(mu) + め(me) + も(mo) 5 8 や(ya) や(ya) + ゆ(yu) + よ(yo) 3 9 ら(ra) ら(ra) + り(ri) + る(ru) + れ(re) + ろ(ro) 5 10 わ(wa) わ(wa) + を(wo) 2 11 ん(n) ん(n) 1 12 が(ga) が(ga) + ぎ(gi) + ぐ(gu) + げ(ge) + ご(go) 5 13 ざ(za) ざ(za) + じ(ji) + ず(zu) + ぜ(ze) + ぞ(zo) 5 14 だ(da) だ(da) + ぢ(dji) + づ(dzu) + で(de) + ど(do) 5 15 – – 0 16 ば(ba) ば(ba) + び(bi) + ぶ(bu) + べ(be) + ぼ(bo) 5 17 ぱ(pa) ぱ(pa) + ぴ(pi) + ぷ(pu) + ぺ(pe) + ぽ(po) 5 18 きゃ(kya) きゃ(kya) + きゅ(kyu) + きょ(kyo) 3 19 ぎゃ(gya) ぎゃ(gya) + ぎゅ(gyu) + ぎょ(gyo) 3 20 しゃ(sha) しゃ(sha) + しゅ(shu) + しょ(sho) 3 21 じゃ(ja) じゃ(ja) + じゅ(ju) + じょ(jo) 3 22 ちゃ(cha) ちゃ(cha) + ちゅ(chu) + ちょ(cho) 3 23 ぢゃ(dya) ぢゃ(dya) + ぢゅ(dyu) + ぢょ(dyo) 3 24 にゃ(nya) にゃ(nya) + にゅ(nyu) + にょ(nyo) 3 25 ひゃ(hya) ひゃ(hya) + ひゅ(hyu) + ひょ(hyo) 3 26 びゃ(bya) びゃ(bya) + びゅ(byu) + びょ(byo) 3 27 ぴゃ(pya) ぴゃ(pya) + ぴゅ(pyu) + ぴょ(pyo) 3 28 みゃ(mya) みゃ(mya) + みゅ(myu) + みょ(myo) 3 29 りゃ(rya) りゃ(rya) + りゅ(ryu) + りょ(ryo) 4 30 – – 0

In level 29, it is also added the study of sokuon, which is a little “tsu” character (っ) that has as its function to double the consonant that comes right after it. Sokuon does not appear in Table II because this table represents hiragana gyos whose characters will be presented alone for the player to type its reading, and once sokuon does not have a single reading – because it depends on the consonant that follows it – it is only used in bosses words, where it can be typed according to its reading in that context. For example, in the word “happi” (はっぴ in hiragana), the first ‘p’ is represented by a sokuon, ergo, its reading in roma-ji is ‘p’; in another example case, there is the word “korokke” (ころっけ in hiragana), where the second ‘k’ is also represented by sokuon, i.e. in this case, sokuon reading in roma-ji is ‘k’. With that said, it is understood that it is not possible to type sokuon’s reading when it is presented alone, since it depends on another characters, and that is why this character is only practiced in words, in this case, Karuchā Ships Invaders’ bosses. Hiragana characters also have different influence/weight depending on the level. The more time a character was presented on the game, the more it will cause damage to the city. For example, in level 2, the character “あ” (a) will cause twice more damage than the character “き” (ki), once (a) was presented since level 1, while (ki) is a new hiragana character, presented for the first time in level 2. And the calculation goes on like this, each level of distance adds 50% more 83 Chapter 5. Development of Karuchā Ships Invaders

damage strength to the ship. For a guarantee that the user will practice more the recent presented characters, there is also a mechanism similar to the one cited above. For each level that passes, the characters presented on previous levels would lose frequency of appearing. Using the same characters on the previous example, in level 2, (a) has 50% less chance to appear than (ki). Also for guarantee that new characters will be practiced, from each two ships, one contains a new character inside, as a way to make sure that new hiragana letters will be presented indeed. Also, the game contains bosses once every three levels, which are pictures related to some of the culture elements cited on Section III, and in the same way, the player has to type the corresponding roma-ji to the picture presented. The bosses are graphically represented by a box with a parachute, with pictures inside that represent a cultural word, and that, if not guided, will bring a lot of damage to the city below. The hiragana and roma-ji for each picture is presented to the player in the stories told by the exchange students, for the player to learn the word associated with a picture, and then be able to type its corresponding roma-ji when the boss falls down, i.e. the boss pictures do not contain tips in hiragana or another alphabet. The percentage of bosses that contain both pictures and characters is 50% for hiragana characters alone and 50% for pictures (bosses). Levels 15 and 30 are different from the others because they are only-boss levels, i.e. no hiragana character alone will fall down, but only cultural pictures will appear on the screen. In level 15, it is possible to appear any boss since level 3 until level 12, plus the new bosses that come as new in the own level. In level 30, the bosses that can appear are all the bosses of the game, since level 3 until level 27, plus the bosses that are first presented in the level 30 itself. Table 9 presents the levels that contain bosses, i.e. stronger ships – boxes with parachutes – that contain, instead of a hiragana character, a picture related to some of the cultural aspects explored in Section III; this Table presents examples of words contained in each boss level, once there are many. The frequency of appearance of bosses – every three levels – demonstrates the dedication Karuchā Ships Invaders has to cultural aspects in the learning of cultural vocabulary through the game. It is important to enhance that each boss level respects the limit gyo, i.e. the cultural elements that are in each level only can contain characters that have already been presented in the game before. Level 3 only contains feeding words, since it is difficult to form words that dovetail in the cultural aspects’ context, with only the first three gyos of hiragana. Table 9 - Some Examples of Bosses of the Game. Cultural aspect (word in hiragana and roma-ji) Level Clothing Feeding Arts & popular culture # すし(sushi) 3 – さけ(sake) – 3 したけ(shitake) 6 うちかけ(uchikake) はし(hashi) とくさつ(tokusatsu) 3 あにめ(anime) きもの(kimono) さしみ(sashimi) からおけ(karaoke) 9 ゆかた(yukata) てまき(temaki) 10 ようかい(youkai) はかま(hakama) のり(nori) かまくら(kamakura) Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 84

まんが(manga) にぎり(nigiri) はおり(haori) うんけい(unkei) 12 ごはん(gohan) 9 げた(geta) えんくう(enkuu) すきやき(sukiyaki) ががく(gagaku) おりがみ(origami) よめそで(yomesode) てりやき(teriyaki) へいあん(heion) 15 ふりそで(furisode) 9 しめじ(shimeji) すどく(sudoku) いろそで(irosode) やくざ(yakuza) いけばな(ikebana) おび(obi) わさび(wasabi) かぶき(kabuki) 18 たび(tabi) やきそば(yakisoba) 10 ぶがく(bugaku) じかたび(jikatabi) てんぷら(tenpura) さむらい(samurai) うきよえ(ukiyoe) いろむじ(iromuji) しょよ(shoyo) 21 じょうもん(joumon) 7 じゅうにひとえ(juunihitoe) しゃり(shari) げいしゃ(geisha) ぞり(zori) ぎゅどん(gyudon) しょどう(shodou) 24 6 おはしょり(ohashori) しゃぶしゃぶ(shabushabu) きょうげん(kyougen) わらし(warashi) わがし(wagashi) しんとう(shintou) 27 6 きゃはん(kyahan) さくらもち(sakuramochi) おんあが (onnagata) にんぎょうじょうる はっぴ(happi) ころっけ(korokke) 30 (ningyoujoururi) 6 しっぱり(hippari) あまなっとう(amanattou) くまどり(kumadori)

The game contains the following requirements for Japanese language learning games identified in a previous chapter (4): i) customizable options, i.e. three difficulty modes (i.e. easy, normal, and hard); ii) immediate feedback (loss of points when a ship hits the city); iii) levels increase, i.e. from 1 to 30; iv) score, i.e. hit points of the player; v) statistics, i.e. top 10 and achievements; and vi) time, i.e. 60 seconds to survive for each level. There are 18 achievements in the game that the player can get, by filling some requirements; these achievements are presented in Table 10, with their symbols (that are highlighted in the achievements screen when got), name and description of requirements to win each one of them. Table 10 - Achievements of the Game. Symbol Name Description Karuchā newbie Finish the first boss level – level 3 – in any difficulty mode

Hiragana newbie Pass the first 5 levels in any difficulty mode

Karuchā beginner Finish the first 3 boss levels in any difficulty mode

Karuchā reckless Did not read any of the stories during a whole game

Karuchā cautious Read all the stories during a whole game

Karuchā on the go Finish level 15 in any difficulty mode

Finish levels with basic characters – until level 11 complete – in any Hiragana on the go difficulty mode Hiragana warrior Finish the game in any difficulty mode

Hiragana lover Finish the game in all the difficulty modes

Karuchā master No loss of hit points in the levels 15 and 30 in any difficulty mode

85 Chapter 5. Development of Karuchā Ships Invaders

Karuchā sensei No loss of hit points in all boss levels in any difficulty mode

No loss of hit points in levels where new gyo are gojuon in any difficulty Gojuon master mode No character with tenten or maru hits the city during the 30 levels in any Handakuten sensei difficulty mode Youon sensei No youon character hits the city during the 30 levels in any difficulty mode

Hiragana master No single character hits the city during the 30 levels in any difficulty mode

Karuchā master No boss hits the city during the 30 levels in any difficulty mode

Karuchā ships invaders No loss hit points during the whole game in any difficulty mode sensei Karuchā and hiragana Played the whole game 10 times in any difficulty mode or summing the addicted modes

It is significant to explain some terms used in the achievements and its importance to hiragana learning. The term “gojuon” designates the most simple hiragana characters, i.e. the ones which do not contain accents. The accents in hiragana can be “tenten” or “maru”. Tenten is a symbol that reminds a quotation mark, put on the top right of a letter to make it sound differently; for example “ka” (か) with a tenten is read “ga” (が). Maru is a small circular symbol used only on the “ha” gyo to make it sound with ‘p’; for example “ha” (は) with a maru is read “pa” (ぱ). The “youon” symbols are combined characters in which the first one is a hiragana that ends on ‘i’, written in normal size, and the second is some character of the “ya” gyo, written smaller, and they create combination of sounds. For example “ki” (き) plus “yo” (よ) gives “kyo” (きょ).

5.4.2 Interface Design and Implementation

Karuchā Ships Invaders was developed to be a game that presents an immersive environment to the player, and this characteristic is meant to be present in all aspects of the game, since the plot, the gameplay and, the most remarkable: interface. Fig. 11 shows Karuchā Ships Invaders main menu interface, with its available options and demonstrating the concept of Japanese ships’ arrival in Brazil.

Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 86

Fig. 11 - Main menu interface of the game. It is important to notice and highlight the presence of key elements that suggest a Brazilian city, in this case, Rio de Janeiro – a worldwide known Brazilian city –, with elements as Christ the Redeemer’s statue – considered one of the modern Seven Wonders of the World –, Maracanã Stadium, favela and the famous Copacabana boardwalk on the beach. The elements that suggest Japan are the ships, which come with a Japanese flag and a hiragana letter and the laser that will guide the ships to the ground, which suggests advanced technology, by whom Japan is widely known. Since every three levels it is presented a set of new cultural words, that are introduced to the player by reporting of the characters (Brazilian exchange students), the level’s choice screen was designed to highlight that (Fig. 12). This screen is divided in the difficulty mode choice area and the level choice area, which presents the unlocked levels corresponding to the selected difficulty mode. Also, the level area is subdivided, and the levels grouped every three, once every three levels, the third is a boss level; still, below each group of three, there is a button that takes to a story, which presents the elements corresponding to the three levels above. There are also two big groups with 15 levels each, indicating that they end up in an only-boss level (levels 15 and 30).

Fig. 12 - Level and difficulty mode interface of the game. 87 Chapter 5. Development of Karuchā Ships Invaders

For each three levels that pass, there is a scenario transition – from morning to evening, from evening to night, and from night to morning again (Fig. 13) – meaning that the experiences of the exchange students are happening in several days and moments of the exploring trip, showing that no matter the hour or the day, there is always something new and interesting to be learned about language and culture of Japan; and this also represents that Japanese language learning is a process that takes a long time and effort to be well done.

Fig. 13 - Scenario transitions of the game, i.e. from/to (a) morning, (b) evening, and (c) night. Fig. 14 shows design elements where characters and words will be put into, being Fig. 14a the element that contains a hiragana inside, and Fig. 14b the element that contains a cultural picture representing a cultural word presented on stories before the level it appears. Following the plot, there are Japanese visitors coming to Brazil, and they are represented by the ships and boxes during the levels. At the end of each level, it is presented the player’s results and score for that level (which is the remaining hit points the player has for that level), and with that, appears a Japanese character which is a sort of pilot of the ship that demonstrates happiness or sadness according to the performance of the player.

Fig. 14 - Ships of the game for (a) hiragana character (b) cultural element. In Fig. 15a there is the character presented when the player passes the level without losing any hit points, and thus, gets the 10% hit points’ bonus; Fig. 15b shows the character presented when the player passes the level but lost some hit point in the process; and Fig. 15c exhibits the character presented when the player could not pass the level and receives the “game over” message.

Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 88

Fig. 15 - Japanese character that represents the player’s performance on a given level, i.e. (a) passed level with bonus (b) passed level (c) game over. Fig. 16 presents a screenshot of the game in execution, explaining the main elements contained on the screen. In Fig. 16 it is possible to observe a ship with a hiragana falling down (Fig. 16a); a box with a parachute bringing a picture of a cultural aspect – in this case, sushi – (Fig. 6b); the area of the city, where ships and boxes will cause damage if not received, represented by a sort of force field around the city – highlighting that this element suggests a Brazilian city (Rio de Janeiro) – (Fig. 16c); the hit points bar, where the green shows how much hit points the player still has (Fig. 16d); indication of the difficult mode being played at the moment (Fig. 16e); the indication of the current level (Fig. 16f); the remaining time the player needs to survive (Fig. 16g); and the element that will shoot the guiding light to the ships and boxes, i.e. the laser machine (Fig. 16h). As a way to differ new characters from previous, new hiragana characters appear with the font color deep gray, whilst characters that have been presented in previous levels come with font color red.

Fig. 16 - Components of the game (a) hit points bar (b) time (c) difficulty mode (d) level of the game (e) ship of hiragana (f) ship of cultural aspect (g) guidance laser ray (h) city (i) laser cannon (j) feedback area of user’s typing. 89 Chapter 5. Development of Karuchā Ships Invaders

Karuchā Ships Invaders CALL game was developed using MOAI platform44 for games development, with Lua programming language45. This development decision was made based on the results of the previous chapter (3), where it was found out the issues of running applications in non native platforms; since MOAI has multiplatform resource, it was chosen for the development of this game, thinking about future distribution for another operational systems.

5.5 Discussion

The first remarkable difference from our game to the others – mainly the games described in Section II – is the exploration of cultural aspects clearly present in the levels of the game – as design items and bosses in certain levels. The use of cultural aspects in the whole environment of the game brings to the player a more immersive way to learn content in a different language, mainly in a case like this one, where there is a Brazilian trying to learn a language from a country – Japan – that is in so many ways different from player’s native land. The game Karuchā Ships Invaders differ from others by a noteworthy feature: it focuses on the hiragana alphabet study. Hiragana is a very important part of Japanese language learning, once it is its basis; with hiragana alphabet it is possible to write every Japanese original word, and that includes all the cultural words present on the game. This alphabet is needed to be known by anyone that aims to read in Japanese language, since it is the most used: for words that do not have kanji, for adjective and verb endings and even for subtitle rare and/or unknown kanji characters. To bring katakana with the idea of the game, it would be necessary to create more different stories of situations that protagonists would go through, with new words to learn, since this alphabet is used mainly to write foreign words, i.e. which do not originate from Japan; with that, more modern Japanese cultural knowledge would be needed, to bring these new words related to experiences that can be lived by foreigners in Japan. Japanese language learning has many particularities, and some of them are directly related to cultural aspects, as the communication means they use daily. When explained out of a context, it can be hard for a Brazilian person to assimilate a topic as language hierarchy, for example, since in Portuguese language we don’t have such thing used often day by day. Another important question about using cultural elements in the design is that it shows Japan to the player, not only in the gameplay – with the relation between hiragana and roma-ji learnt by playing – but in the whole process of being in contact with the game itself – through images and sounds, for example. These elements bring the player to know more about Japan whilst learning the language of this

44 http://www.getmoai.com. 45 http://www.lua.org. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 90

country, and the harmony between learning and design is pleasant, helping and encouraging even more the learning process. It is also important to notice the influence of Brazilian culture itself on the development of this kind of game. Since it is planned and designed by/for Brazilians, it is expected that the life experiences of the programmers and designers, and the cultural environment where the game was developed, can influence on the final result too. Still, this is a game developed thinking first in Brazilian students of Japanese, so it is necessary to think over the knowledge and experiences of the players and how to take them and, in an effective way, bring them to the game and the learning process too. That is why the game mixes Brazilian and Japanese elements, since the plot, where the two Brazilian exchange students go to Japan and inform their friends in Brazil about what they learn every day; to the arrival of Japanese ships in a Brazilian city, that will be received and guided for these friends with the knowledge they got from the students. Another particularity of our game is to bring out several game requirements, as choice of difficulty, for attending since the beginner willing to learn hiragana, a student that wants to learn more of the Japanese popular vocabulary with the cultural aspects, until a more expert student that just wants to practice knowledge; several levels that increase hardship for the player, avoiding the game to become boring and keeping it interesting for the player to get motivation to go through all the levels; a top 10 with the names and scores of the players who went better, being also a motivation trigger for the player to get best scores and play more, thus learning more of hiragana and Japanese culture. Karuchā Ships Invaders is available in PT-BR language, once it is focused on the Brazilian Japanese learners, and this feature allows a lot more Brazilian users to access this kind of resource, because with the games currently available for this purpose, the student was forced or needed to know the English language to be an intermediate between Portuguese and Japanese. This kind of situation is harmful for the learning process, whereas each language has its own particularities that are already difficulty to perceive from one to another; once a third different idiom is brought into the process, it becomes a lot easier to lose details that can compromise a good knowledge of a given language.

5.6 Conclusion

This chapter presented a CALL game with the purpose of helping teaching Japanese language, introducing the hiragana alphabet; developed taking into account aspects of the Japanese culture – included in the gameplay as the biggest enemies (bosses) –, also introduced in this chapter. It was also performed a study of Japanese cultural features to include it in the game design, creating a more immersive environment to the user, improving the learning experience. 91 Chapter 5. Development of Karuchā Ships Invaders

As future steps of the research, we intend to develop katakana mode, with all of its symbols and combinations as hiragana, and brand new cultural words that are written with this alphabet. In a super advanced mode, it is planned to add kanji words in the bosses. Also, we point out the accomplishment of tests with Japanese language students to collect data related to learning and interface/interaction evaluation of the game.

5.6.1 Decision Making

The decision made after the accomplishment of this step of the research is to perform an initial interface evaluation, in order to discover what are the points in which the tool needs to be improved before performing an user evaluation. Since we intend to evaluate other aspects, such as motivational, it is not interesting that the game be presented to prospective users with basic interface failures. To accomplish that, we intend to search for evaluation methods that are focused in this kind of tool, and if there is no such method, the idea is to propose a new one, based on an already used method in the evaluation of interface aspects in general.

Chapter 6 6 Interface Evaluation of Karuchā Ships Invaders

6.1 Introduction

Interface evaluation of computer systems in general include usability and accessibility as main aspects, and when it comes to the field of computer games, another aspect arises, which is playability. Playability measures aspects of gameplay, such as motivation, satisfaction, immersion and challenge. To accomplish these evaluations, manual, semi-automatic and automatic methods can be used; each one of them can detect different interface issues. The interface evaluation brings for the developers several questions and points of view about a system, which they can use to improve it and solve problems for the end user. The accomplishment of an interface evaluation is fundamental to analyze the Human-Computer Interaction (HCI) aspects of a given system. In this context, Ivory and Hearst [76] present the state of the art about automating methods of usability evaluation of user interfaces, exhibiting a taxonomy that highlights the role of automation in this context and to compare several methods. 132 methods are evaluated, applied between WIMP and web interfaces. They highlight that user evaluation is fundamental, but can be increased and complemented by automated methods, and also contribute for lower costs and time of the evaluation as a whole. Følstad et al. [71] present a survey on practical usability evaluation, using a methodology that includes observation and qualitative interviews. The survey respondents were 224 usability practitioners with experience, who answered questions about their experiences in usability evaluations they have participated. The conclusion is the evaluation techniques are often adapted by the practitioners, for the given context. 93 Chapter 6. Interface Evaluation of Karuchā Ships Invaders

Hilbert and Redmiles [75] performed a survey about user interface events, focused on computer-aided techniques, in which they present a framework to help on approaches’ categorization and comparison. The authors concluded that computer-aided techniques for evaluation of interface can bring out the overall problems that can be detected by an automatic method, but lacks points that can only be perceived by the analysis of a human being, so the techniques are complementary. Grossman et al. [73] perform a survey about learnability, which is an important aspect of usability, that represents how easy is to the target users to learn how to use a system. It is compared a traditional think-aloud method to a question-suggestion method, using 10 undergraduate architecture students, volunteers, to evaluate AutoCAD’s learnability. The results made the authors develop a new evaluation protocol to identify learnability issues, and they conclude that the methodologies can be used together. Yuan et al. [94] accomplish a survey about accessibility of games. The impairments are classified in visual, hearing, motor and cognitive, and can lead to two types of barriers: (i) critical, when the person cannot play a game because of a disability; (ii) and non-critical, when the person is able to play, but with reduced gaming experience. To reduce these barriers, it is pointed the use of several kinds of stimuli for an activity, as visual, audible, and haptic. The authors provide manifold detail about each disability in various kinds of game; still, provide directions in development and evaluation of games with characteristics of accessibility that embraces several kinds of audience. The growing use of technology to enhance learning experiences grows increasingly in the field of language learning, these educational applications are called Technology Enhanced Language Learning (TELL) in literature. TELL has two main branches, which are Computer Assisted Language Learning (CALL) and Mobile Assisted Language Learning (MALL), and in both are used applications for a given operational system, that can be web and/or standalone, that will help the user to learn a target language. CALL and MALL applications can be tutor systems, questionnaires, multi-user sharing experience; and a rising type of application is game. In this context, in previous steps of the research, we identified Japanese CALL tools from academic and practical environments, including desktop, web and mobile systems. We identified 23 academic works, divided into desktop, mobile, web, haptic, and augmented and virtual reality applications [35]. In the practical field were analyzed 30 tools for web [43] and 64 Android apps [62]. The results obtained led to the development of a CALL game to teach Japanese language [80]. The present chapter accomplishes a literature review about interface evaluation focused on educational computer games; with the goal to perform an interface evaluation of our own CALL game developed early. The game is called Karuchā Ships Invaders and it is meant to teach the hiragana alphabet and Japanese cultural aspects for students of Japanese language. This chapter is organized as follows: Subsection 6.2 presents the Karuchā Ships Invaders CALL game; the Subsection 6.3 exhibits the literature review; after that, Subsection 6.4 presents Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 94

the study of the interface evaluation for the game; Subsection 6.5 brings the discussion; and, finally, Subsection 6.6 concludes the chapter.

6.2 Karuchā Ships Invaders CALL Game

The Japanese language is a very challenging idiom to learn once it has three main alphabets that are all different from the Roman alphabet used in most languages around the world. There are kanji, ideograms that were originally Chinese but were imported and adapted to Japanese language; they represent ideas and one kanji alone can have several pronunciations. The other two Japanese alphabets are hiragana and katakana (which together are called kana), they are different from kanji since they are syllabic alphabets; the former is the most basic, used for writing words that do not have a kanji, also for adjectives and verb endings, and as a subtitle for rare or unknown kanji (the subtitle is called furigana). Hiragana is the first taught alphabet to anyone learning Japanese language, in or outside Japan. Katakana is an alphabet similar to hiragana in its basis, but the symbols are all different and it is used to adapt foreign words to the Japanese idiom, and in some cases for highlighting a certain word or onomatopoeia. In Japan, they also use the Roman alphabet – which is used in most languages –, but barely; for example, in acronyms. In this context, Karuchā Ships Invaders is a CALL game46 developed by our research team, and presented in details in [13], with the goal of supporting Japanese language learning, by means of presenting the most basic Japanese alphabet – i.e. hiragana – plus some cultural words related to Japanese culture. The concept of the game is a story where Japanese ships are heading to Brazil – represented by a city with Brazilian elements on the game screen – and need help to land; this help has to come from Brazilians that know the basics of the Japanese language, to pass the commands to the ships. The gameplay is space invaders-like, where the ships act like “invaders”. Each ship brings a Japanese letter on it, or a cultural Japanese word written with the basic alphabet; and the player has to type on the keyboard, the corresponding pronunciation using the Roman alphabet, and then press the Enter key or backspace. If the type is correct, a laser will be shot and embrace the ship, helping it to land safely, though, if a ship is not guided – the player does not type its pronounce –, it will fall down and cause damage to the city below. Fig. 17 shows three screenshots of the gameplay: Fig. 17 (a) show the level choice screen, Fig. 17 (b) and (c) shows gameplay screens, Fig. 17 (d) shows the end of a won level, Fig. 17 (e) shows a context-help menu, and Fig. 17 (f) shows the end of a failed level.

46 Available for free download at http://www.karucha.pairg.dimap.ufrn.br. 95 Chapter 6. Interface Evaluation of Karuchā Ships Invaders

(a) (b)

(c) (d)

(e) (f) Fig. 17 - Different moments of the Karuchā Ships Invaders CALL game. To present/teach new content to the player, the game exhibits stories protagonized by two Brazilians that went to Japan to learn about its culture; the stories are reports to their friends in Brazil, to help them be able to guide the ships that come to land. The game affords three difficulty modes – i.e. easy, normal, and hard – each one increasing velocity to the ships’ fall. Still, each difficulty mode contains 30 levels to be passed, with increasing number of hiragana characters and cultural words to be learned. Hiragana is presented in alphabetical order and is exhibited in all levels but 15 and 30, which are levels that only contain cultural words called the bosses of the game. Also, every three levels, the third combines hiragana characters with cultural words, or bosses. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 96

The Karuchā Ships Invaders game was developed to provide an immersive environment for the users – with a focus, but not limited to, on Brazilian students of the Japanese language –, presenting the Brazilian city that receives Japanese visitors; in Brazil side there is the city elements that reminds of Brazil, like, for example, the statue of Christ the Redeemer, Maracanã Stadium, favela and the famous Copacabana boardwalk on the beach. Japan is represented on the ships, which bring a Japanese flag in their front, plus the letters themselves, that are unique of the Japanese writing system. Multimedia resources also contribute to the immersive nature of the game, using, for example, sounds and images that remind of both Japan and Brazil.

6.3 Interface Evaluation of Educational Computer Games

The literature review is a fundamental part of this chapter, because it is from this that analysis of methods, with applications and results, is made. This overview presented next enables getting knowledge about evaluation in the context of educational games, taking into account several aspects of interface to be evaluated. Once this is a very specific field, there is not a large amount of works related; still, this does not intend to be an exhaustive list of related works, but a selection of the main research and techniques currently used in this context. The search methodology includes searching papers in main online scientific libraries and bibliographic databases, using certain keywords and concatenations of them. The used libraries are ACM DL, IEEE Xplore, Springer, Elsevier, Scielo, Scopus, ISI Web of Knowledge, and also in Google Scholar. The keywords searched are a permutation of the following words (presented in this chapter in alphabetical order): “accessibility”, “case”, “educational”, “evaluation”, “game”, “interface”, “literature”, “playability”, “review”, “state of the art”, “study”, “survey”, and “usability”. After the search was accomplished, works were selected to compose this review, based first on the abstracts and, second, a nimble reading of them. The selected works were then read carefully, analyzed and classified according to methodology, aspects and requirement of users in the process. The chosen works are presented next. Omar and Jaafar [88] aimed to reach a heuristic method to evaluate educational games, with focus in Malaysian ones; and with what it would be necessary a few persons to evaluate, including designers, developers and usability professionals. They use heuristics based on Nielsen’s, relating them to the context of educational games, for evaluation of playability. Another work by the same authors [86] brings a set of heuristics and a framework of five steps to apply them in an interface evaluation using prospective and end users, with heuristics focused on educational games that were divided into five aspects of educational games, i.e. interface, educational/pedagogical, content, multimedia, and playability. In a later work of these authors [89], Omar et al. improved the previous work presenting a tool – AHP_HeGES – to facilitate 97 Chapter 6. Interface Evaluation of Karuchā Ships Invaders

this kind of evaluation during the game’s development process. The methods are not validated in those works. Desurvire et al. [69] bring a set of heuristics to evaluate playability of video, computer and board games, based in other heuristics from the literature, and to be made by an expert evaluator; they called the method HEP. Their method is applied in the work with the development of a game which is used to evaluate the method. They also made a user centered evaluation, using four prospective users, in order to compare its results with the HEP method and concluded that HEP found a considerable higher number of issues over the user centric approach, despite the latter found more specific issues to the game. Liao and Shen [79] present the interface evaluation of a computer game-based learning material, called Crazy Machines 2, using heuristics and user centric approaches to determine if the game was developed under usability principles. The heuristics used are present in a work referred by the authors; and the user centric evaluation was made with five research students from courses as education and communication technologies. Korhonen et al. [77] accomplish a comparison work between two playability heuristic sets. The evaluation was performed with eight persons from game industry and academy, divided in four pairs. The results of the study point out that the use of heuristics guided the evaluators well, including during the development phase; a lower point of heuristics, according to the evaluators, is that since there are a lot of kinds of games, it is difficult to make heuristics that work for all of them; on the other side, too much specific heuristics may restrict their use. Another interesting finding is that it is hard for the evaluators to have a real game experience during the evaluations, once they have to be alert to the playability problems that may occur. The authors concluded that this is a valid method, but that needs improvement for the games area. Pinelle et al. [91] develops a new set of heuristics to evaluate usability of game interfaces. They base their work on Desurvire’s [69] and claim that the HEP heuristics have limited coverage of usability issues; then, they presented a set of heuristics focused on usability, made by the authors. To do that, the authors analyzed a list of 108 computer game reviews, and once the problems were identified and classified, it would be developed heuristics that would be the opposite of the given problem. They reached a number of 12 categories of problems, and developed a list of ten heuristics, highlighting that they are based on Nielsen’s. The evaluators were five videogame players with experience in usability evaluation. Barcelos et al. [68] propose another new set of heuristics for evaluation of digital games, and validate them comparing to another existing set. Their set is a compilation of some others, including Pinelle’s [91], and what they aimed to do was to create a more short and direct set of heuristics, that could be as efficient as the larger ones in which what they based. For the evaluation, were chosen the games Earth 2160 and Outlive, both from the category Real Time Strategy of games. The evaluators were 34 HCI students of a upper course of digital games; they were divided in groups in which some of them received the list of 18 heuristics developed by the Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 98

authors, and others received the 40 Federoff’s [70] heuristics. They conclude, by the results of the evaluation, that the performance of both sets is very similar. Grace et al. [72] provide an initial evaluation of a language learning game for Mandarin. The evaluation was made with 21 potential users of the game, called Polyglot Cubed. By means of questionnaires, the authors analyzed the response of the users and were able to detect some interface issues with the game, as problems with selecting objects while playing and seeing pictographic representations of words. Serrano et al. [92] accomplish an evaluation of an educational game through observation of users’ interaction. While the players do play the game, data is collected automatically and analyzed to extract relevant issues. They proposed a framework and performed a case study with a basic math game to validate it. Lazareck et al. [78] performed an evaluation of an educational game for 13-15 year olds, which teaches about biotics and hygiene. The evaluation was made with 129 pupils of three different UK schools, in a user centric approach that used a questionnaire the users answered after playing the game. The questions addressed mainly to usability and playability, and the answers were given in varied ways, since yes/no ultimo comments they wanted to make about the game, and then the responses were analyzed. Hersh and Leporini [74] performed a study about accessibility in educational games for disabled students. The study was applied on disabled students and people that are close to them, as family and educators, by means of a questionnaire. The results of the study gave the authors answers for them to develop guidelines and recommendations about the development of this kind of game for these target users. The guidelines, beyond accessibility, also evaluate pedagogical aspects and playability/usability. Park and Kim [90] present a study about accessibility for disabled people, in serious games for mobile environment. The authors suggest guidelines for development of this kind of game taking into account the following aspects: definition of game goals; planning game scenario; planning game components; planning rewards; planning training progress; and planning of training effects and achievement. Each one of these aspects has a guideline associated. Torrente et al. [93] bring a study about accessibility of educational games for blind students. They developed three different eyes-free interfaces, applied on a point-and-click game. The cyclical navigation system provides a two-level interaction using right and left arrows, and buttons “enter” and “escape”; in which the elements are selected and the options are passed through sounds for the player. The second, called sonar, uses the mouse and sounds to indicate that the cursor is near an object. And the third is a natural language commands interface, with typed commands. It was made an evaluation of the three interfaces analyzing them according to usability, engagement and additional cost. They also mention a pilot study with ten blind students, though it is not described results about this study. Table 11 summarizes the works of literature previously presented. This table presents the column names: “Based on”, which indicates if the method is based on (H)euristics, User Centric (UC), (An)alysis, and/or (G)uidelines; “Focus” that indicates the evaluation focus, i.e. (U)sability, 99 Chapter 6. Interface Evaluation of Karuchā Ships Invaders

(A)ccessibility, (Pl)ayability, (I)nterface, (Pe)dagogical, (S)ocial, (M)ultimedia, (C)ontent, (St)ory, (R)ules, and/or (Me)chanics; “Applied” and “Needs User” that indicates, respectively, if the method was applied and if the method needs end users during the evaluation process, i.e. (Y)es or (N)o. Table 11 - Summary of the works of literature review. Evaluation Work Based on Focus Applied End Users [88] H Pl N N [86] H I, Pe, C, Pl & M N Y [89] H & UC I, Pe, C, Pl & M N Y [69] H & UC Pl, St, Me & U Y N [79] H & UC I, R, St & S Y Y [77] H Pl Y N [91] H U Y Y [68] H Pl Y N [72] UC U Y Y [92] An U & Pe Y Y [78] UC U & Pl Y Y [74] G U, A, Pl & Pe Y Y [90] G A N N [93] An A Y Y

It is possible to notice a trend of heuristic evaluation followed by user centric evaluations. The heuristic evaluation has the advantage of being flexible and can also be applied with prospective and end users. Related to the focus, the most evaluated are playability, usability and pedagogical aspects. The works that evaluate accessibility are specific and use techniques as guidelines and comparative analysis. The majority of works apply their techniques to validate them practically. Some of the techniques presented are just exhibited and explained, but not applied in the same work.

6.4 Karuchā Ships Invaders Study

In this Subsection, we first present the methodology used to evaluate multiple aspects of interface of Karuchā Ships Invaders CALL game, and then another subsection presents the results obtained in this study.

6.4.1 Methodology

The methodology chosen to be applied in this study is, initially, heuristic based, the developers being evaluators, for detecting overall interface issues, based on the work of Omar and Jaafar [86]. The choice of this work in particular was made because it has a set of heuristics focused on evaluation of educational computer games, which is our focus too. Still, this work [86] is a result Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 100

of several research [88][89][87][81][83][82][85][84] focused on this subject-matter. The work [86] brings the set of heuristics and also provides a framework with five steps to apply, represented in Fig. 18.

Fig. 18 - Five steps of the evaluation method. The first step consists in developing a questionnaire based on the heuristics that suits the context of the game to be evaluated. After that, it is made an identification of possible evaluators, between experts, prospective and end users. The third step plants the way the evaluation will be conducted, as location, time and form of presentation of questionnaire. The fourth step designs the evaluation as a whole, determining which will be made in the available time and the order to be followed. The fifth step is related to the analysis after the evaluation and dictates what will be made with the results. The steps conducted in this study are presented in this section. The heuristics used in the expert evaluation – Playability Heuristics for Educational Games (PHEG) – are divided in five main aspects of an educational game, i.e. (i) Interface, which includes factors as interactivity, navigation, design and consistency (UI has ten heuristics); (ii) Educational/Pedagogical, which evaluate goal and objective, challenge, feedback and player control (ED has ten heuristics); (iii) Content, which contain aspects related to the educational material provided (CO has eight heuristics); (iv) Multimedia, which addresses to the use of text, audio, animations and all kinds of media, if they are appropriate (MM has eight heuristics); and (v) Playability, which evaluates questions such as balance, pace, control of the player over the game, and adequacy of levels (PL has seven heuristics), in a total number of 43 heuristics. The PHEG evaluation can be performed by experts, or by prospective/end users. For the evaluation, we developed a set of questions based on the basis heuristics of five aspects previously described, for guiding the evaluators in the process. The method requires two or more questions per aspect, but it is not mandatory to develop questions over all the heuristics. The questions of this study were developed based on the heuristics that fit the context and needs of Karuchā Ships Invaders CALL game. Table 12 presents the heuristics used as basis for the development of the questions (cf. [86]); in this table we do not present all 43 heuristics, only the ones which we used to develop questions for our research. 101 Chapter 6. Interface Evaluation of Karuchā Ships Invaders

Table 12 - Basis heuristics for the development of the questions. Heuristic Aspect Code Brief Description

UI1 Uses aesthetic and minimalist design

UI3 The uses of space, color and text are according to the principles of screen design UI4 The uses of text, color and font follow the principles of readability

Interface UI6 The interactivity of the game is suitable to learners level UI9 Provide specific and self-identified key for specific task (e.g. exit, glossary, main, and objective)

ED1 Clear goal and learning objectives ED2 The activities are interesting and engaging

/ / ED4 Can be used as self-directed learning tools

ED9 Offers the ability to select the level of difficulty in games Educational Pedagogical ED10 Ability to work in their own pace

CO1 Reliable content with correct flow

CO3 Navigation of content is easy and accurate CO4 Supporting materials are sufficient and relevant

Content CO5 Materials are interesting and engaging CO6 Players able to understand the learning goal

MM2 Combination of multimedia elements are adequate

MM3 The presentation of multimedia elements are well manage MM4 Suitability of multimedia elements for specific use MM6 The use of multimedia elements support meaningfully the text provided

Multimedia The quality of multimedia elements (e.g. text, image, animation, video, and sound) used is MM7 acceptable

PL3 The pace of the game are in balance

PL4 Players able to control the game ty PL5 Progress of the game can be seen at anytime Playabili PL7 Challenge is adequate, i.e. not too easy and not too difficult

Table 13 presents the questions developed specially for the evaluation of Karuchā Ships Invaders CALL game, which represents the first step of the framework. Table 13 - Questions defined to assess multiple interface aspects of Karuchā Ships Invaders. Basis Heuristic Aspect Question (Table 12) UI1 Q1. Graphical elements are consistent and adequate?

UI3 Q2. Screen space is well distributed (each element has enough space)?

Q3. Font size for roma-ji is adequate? UI4 Q4. Font size for Japanese characters is adequate?

Interface UI6 Q5. Interaction mode is adequate to the audience? Q6. Access buttons for different functions and features are easily identified? UI9 Q7. Menu division is adequate? ED1 Q8. Learning goals are clear? ED2 Q9. Gameplay is incentive and engaging?

Q10. Is it possible to obtain new considerable knowledge about Japanese language by ED4 just playing the game?

ogical Q11. Does the game provide choice of difficulty level? ED9 Q12. The difficulty levels are adequate, i.e. too easy/hard, or discrepant? Education/Pedag ED10 Q13. The game suits to different learning paces of the students? CO1 Q14. The content is presented in gradual and coherent way?

CO3 Q15. Is it easy to access support content, i.e. stories? Q16. The educative stories bring the content in a comprehensive way? CO4 Q17. The educative stories do provide enough background for the player to win the Content game? CO5 Q18. Is it interesting/engaging to learn through the stories? Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 102

CO6 Q19. Is it clear for the player the content to be learned along the game? Q20. The multimedia elements (e.g. sounds, images, and text) are in the context of the MM2 game? Q21. The presentation of multimedia content follows an adequate flow? (Is there too MM3

much or a lack of multimedia)

a Q22. The use of the following multimedia elements is adequate? a. Images in the stories b. Text in the stories MM4 c. Ambient music

Multimedi d. Sounds (e.g. laser, shot, and explosion) e. Images in level transitions Q23. The multimedia elements are significant for the gameplay and learning (or are MM6 dispensable)? MM7 Q24. The quality of multimedia elements is acceptable? PL3 Q25. The game flow is adequate? PL4 Q26. The controls of the game are adequate (or are tiring, hard, non-intuitive)? Q27. Is it easy for the player to perceive or access the current status in the game at any PL5 moment?

Playability Q28. The challenge level is adequate to the audience, i.e. not too easy and not too PL7 difficult?

6.4.2 Results

The evaluation was made using HCI experts as respondents for the questions developed about Karuchā Ships Invaders CALL game. The answers were given subjectively, since the method used is an adaptation for Omar and Jaafar [86]. It was chosen to analyze the data beyond objective numbers as answer. Table 14 presents the answers for the 28 heuristic based questions previously defined (Table 13). Table 14 - Answers for the heuristic based questions. Aspect Answer Q1. The graphical elements are consistent together, and also are in the context of the learning proposal of the game. Q2. The game screen could be increased in height, for the elements to have more space to fall. Though this change could require other adjustments about resolution. Q3. The roma-ji fonts along the game are in an acceptable size, but the game should provide personalization for

this feature, for possible players with sight issues. Q4. The characters inside the ships are in a size sufficient enough to be seen clearly. Though, it is interesting to give the player the opportunity to personalize this feature.

Interface Q5. Considering the audience as learners with abilities to see the elements on the screen and type commands on the keyboard, the interaction mode is adequate. Q6. In the menu screen and inside the menus, the elements are identifiable, but when the game is on, there is no button that indicates pause, menu, configurations of sound or screen. These elements should be represented by icons while the game is being played, for the player to have more control over the game. Q7. Yes, the menu options are well divided. 103 Chapter 6. Interface Evaluation of Karuchā Ships Invaders

Q8. There is not a presentation of the game in the beginning. The help menu provides some information, but it would be interesting to have some kind of initial animation to present the educational and fun goals of the game. Q9. The gameplay is invaders-like, so it can be engaging for those who like this kind of game. Also can be incentive the fact that the player is learning a new language while playing the game. Q10. Considering the player with no knowledge at all about Japanese language, s/he will get considerable knowledge (an entire alphabet and vocabulary – words related to the culture of the country). For the ones who play Karuchā Ships Invaders for training hiragana skills, there are still the cultural words to be learned. Q11. Yes, the game provides three difficulty modes.

Educational/Pedagogical Q12. The difficulty modes are just different in velocity. Considering that, they are not discrepant. But related to content, there is no increasing based on the difficulty. Q13. The game provides three different difficulty levels. To suit all paces, it is desirable that the game provide some kinds of personalization, which it does not.

Q14. Yes. The hiragana content is presented in alphabetical order, and the cultural words are presented just when all the characters that compound it have been presented before. The amount of content by level is also acceptable. Q15. The stories become available as the player passes through the levels (one story for each three levels). Once available, the player has the option to read it as many times as needed, in the level choice screen.

Content Q16. Considering the audience is a native English or Portuguese speaker. The content is well divided and exhibited for the player. Q17. The stories present all the content that will be charged in the game, so it provides all the knowledge the player needs to reach the end of Karuchā Ships Invaders. Q18. The stories are displayed in text and images. Maybe animations could be more engaging and less tiring. Q19. The content is only viewed as a whole when the player gets to the end of the game (30 levels).

Q20. In general, the multimedia elements follow the context of the game. The music and sounds could be more related to the theme. Q21. The multimedia is in good amount. Q22. a. The images in the stories are fundamental to teach the cultural words, once they are the same images that will appear in the game for the player to recognize. b. The text in the stories is adequate and fundamental. c. The ambient music could be more related to Brazil and Japan at the same time.

Multimedia d. The sounds for laser, shot and explosion are common sounds that can be adequate, but they could remind more of the theme (Brazil – Japan) e. The Japanese pilot is adequate to the context. Q23. The use of educational stories is more acceptable for the player when it presents not only plain text, but images and animations; the multimedia elements should not be considered dispensable in this case. Plus, the gameplay becomes more joyful when the game provides an immersive environment related to the theme, what happens in Karuchā Ships Invaders. Q24. The elements of multimedia are in good quality. Q25. The presentation of elements follows a logic order and the number of levels corresponds to the content that is well distributed between each level. The difficulty grows increasingly including the cultural words that can be formed only by the characters that have already been presented. Q26. Considering a desktop environment, the controls feel adequate, once it focuses on the standard keyboard. The controls are intuitive, since the player just has to type the roma-ji to the symbols and words presented.

Playability Q27. The time, difficulty mode and level are clear on the screen; the hit points bar does not indicate a fix number or the amount of loss in a crash. The total points are also not exhibited. Q28. Considering the audience as learners of hiragana that can be from knowing nothing to be conversant to the alphabet; for the former, it can be considered little hard if the player chooses not to read the educative stories that are presented. The cultural words can be a plus challenge for those who only learnt the alphabet itself. The possibility of choice of difficulty mode embraces most of the audience needs.

The results provided, i.e., the answers by the experts (Table 14), are the data to be analyzed regarding Karuchā Ships Invaders’ interface. In the fourth step of the process, it was decided to apply the questionnaire, then analyze the responses, then generate quantitative results based on these responses, leading to detection of specific number of problems and its corresponding severity rating. Table 15 presents the severity rating for each one of the five aspects evaluated; Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 104

beside the severity ratings, there is an indication of what question detected the issue. This table is a representation of what was decided to be made in the fifth step, based on the framework. Table 15 - Severity rating of problems found. Problems Aspect Severity rating Number 4 3 2 1 0 3 1 Interface 4 - - - (Q3, Q4 & Q6) (Q2) Educational/ 1 1 2 - - - Pedagogical (Q13) (Q8) 1 Content 1 - - - - (Q18) 1 Multimedia 1 - - - - (Q20 /Q22c /Q22d) 1 Playability 1 - - - - (Q27)

It is interesting to highlight that no problem with severity rating 3 or 4 was found in this evaluation. This may have with the fact that Karuchā Ships Invaders was developed over considerable previous research [35,43,62], providing a better sight of what was desirable to be developed. A discussion about the results is presented in the next section.

6.5 Discussion

This chapter brings a noteworthy contribution to the field of evaluation of game interfaces in general, once it brings a literature review that can guide researchers that aim to accomplish this kind of evaluation. Still, the evaluation that was performed for Karuchā Ships Invaders also shows that methods can be adapted for the needs and context of each game, focusing in educational computer games. The framework and heuristics used are adaptive and can be applied for experts or for prospective/end users, depending on the questions developed based on these heuristics. The answers for the questionnaire also raise discussions that follow, about each one of the five analyzed aspects of usability. The interface analysis shows that there are some issues such as lack of personalization for features as font size, and lack of buttons and configuration elements at the time a level is being played; this analysis also raises questions about accessibility of the game, once there are no customizable options to support players with special needs. For the educational/pedagogical aspect, the issues are about the difficulty modes, which do not increase content to be learned in each mode. This question can be answered analyzing the sort of the game, which is “invaders”; this kind of game has the characteristic of increasing speed as the difficulty increases, and related to new content, it is represented by the 30 levels, where each one of them brings new characters or words for the player to learn. Besides, there is a discussion about the learning goal being clear to the player, since the content is not provided at 105 Chapter 6. Interface Evaluation of Karuchā Ships Invaders

once in the beginning of the game; the question is if it is really necessary, or in and educational and fun point of view it is interesting to the player to discover new content along the game. The purpose of the locked content is to make the game more engaging and interesting. The analysis of content perceives that the presentation of the stories with only text and pictures may be exhaustive for the player. The suggestion, to make the learning more gamified, is to add animations to the stories; the animations are less tiring to the player and more interesting. After all the game is a new way to learn, with the bid to be different from traditional learning, in which reading is a noted constant, and perhaps should be avoided in this approach. Multimedia aspect brings attention to the sound and music aspect, which are the less related to the context of the entire game. The ambient music reminds only Japan, whilst the other sounds are way too common; the latter is not a big issue, but fixing it could render the game a more interesting and immersive learning environment. The analysis of playability aspect, as multimedia, was the one that showed the fewest issues of all; the main question was about the possibility of the player to get access to the current statistics of the game at any time, which was found that there is a lack of this possibility. In this case, this lack can be seen as a gamification characteristic, whereas the player is having fun with the game without the preoccupation of looking at statistics when a level is being played; this feature takes this load away from the player, who is concerned about protecting the city and guiding the ships to land, instead of a learning status. About the technique used to evaluate the interface of Karuchā Ships Invaders CALL game, it was useful to evaluate the five aspects, but there is a lack it does not embrace. The emotional and affective aspects of the interface cannot be evaluated by this set of heuristics. In the context of educational computer games, as seen in this literature review, this aspect is not contemplated in general evaluations.

6.6 Conclusion

This step of the research accomplished a literature review about interface evaluation of educational games, followed by a study that applied game interface evaluation techniques in the Karuchā Ships Invaders CALL game. The performed evaluation embraced the points interface, educational/pedagogical, content, multimedia and playability, and showed that the game interface is in general, adequate to its purpose. Still, it can be concluded that the method used is useful, but does not embrace certain aspects, such as the player’s emotional response to the game. As future work, it is intended to do a hedonic evaluation of the game. Also, it is pointed to improve the interface of the Karuchā Ships Invaders CALL game based on the results of this study, plus do an evaluation that focuses on learning value of the game with end users.

Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 106

6.6.1 Decision Making

This step of the research showed that, fixed some interface issues, the game is in conditions to pass through a user evaluation. The next step we decided to take is to search for methods that can evaluate aspects as motivational, with focus in educational computational tools, in order to choose the one who fits better. In case that there is no such method, the same line of this chapter will be followed: search for methods that are closer to our research context and adapt it, proposing an extension or a new method.

Chapter 7 7 Evaluation of Motivational Aspects47

7.1 Introduction

The use of technology to support several activities is a reality nowadays. In this context we have, among others, the use of technological tools that support the learning of manifold subjects. In previous chapters (2, 3 and 4) we accomplished some research about this theme, evaluating several of these tools, from academic and practical field, and perceiving its platforms, classifications and common requirements/features; including a chapter focused on educational media applied specifically on these kinds of tools. In this context, this chapter intends to study and perform a method proposal for evaluation of the motivation a technological educational tool causes over a student’s will in learning a given subject through it. This evaluation is in the middle between interface and learning evaluations, being an axis that leads to learning, whereas learning is not necessarily pleasant; so before performing a substantial and long learning evaluation, in this technological context, it is important to inquire if the given tool is capable to hold the student in the process in a pleasant way, i.e. if it offers an environment that induces the student to keep studying the given subject. As a guide, we use ARCS Model [98], which is a method for evaluation of motivation in traditional courses, but an example of application of this method in a computational system is Huang et al. [97], which uses ARCS model as basis to evaluate a tutoring system that teaches MATLAB syntactical structures; it also uses a method of analysis based on ARCS, the IMMS (Instructional Materials Motivational Survey) [99] that provides a standard scale to be used for answering the questions; in this study, the result’s analysis is made crossing and comparing profiles with respective answers.

47 Este capítulo é uma versão revisada do trabalho [108], já publicado e apresentado. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 108

In this chapter, we propose a heuristic-based method to perform an evaluation of motivational aspects of educational technologies. The proposed method uses heuristics as a basis and involves experiences with end users of the evaluated tool. We describe in details every steps of the evaluation method, since the development of the necessary formularies, identification of users, the moment of the experience and the data analysis after that; with the objective of help other researchers to apply the method in their own research contexts. This chapter is organized as follows: subsection 7.2, we present an understanding of motivation and introduce some basic terms and concepts to help understanding the chapter as a whole; following, in subsection 7.3 we detail the method proposal developed; after that, subsection 7.4 we present the discussion section; and finally, in subsection 7.5, the conclusion is presented.

7.2 Motivation

This section presents some basic terms and concepts – i.e. interface, heuristic and, mainly, motivation – adopted along this work, aiming to better specify the intentions of the proposed method and to clarify the meaning of the used terms. Motivation is a term used in many contexts, therefore it has different definitions and/or understandings for each one of them. In psychology, motivation refers to a condition in the organism that influences the direction of behavior. There are several theories about motivation, for example Maslow’s (pyramid of needs), Murray’s (needs and motives) and Skinner’s (conditioning) theories. Motivation is also related to hedonism, which embraces the pleasure and satisfaction of the human being. The term is defined, in the Cambridge Dictionary, as “willingness to do something, or something that causes such willingness”. Still, motivation can be classified in two kinds, i.e. intrinsic and extrinsic. The intrinsic motivation is when the person does some activity for his own sake, because he enjoys it, and not searching for an external reward. The extrinsic motivation is the opposite, the individual accomplishes a given task because of a reward to be earned, or a punishment to be avoided [103][104]. Through these definitions, we understand that the motivational aspects of an educational tool are the ones that have the power to keep the interest of the student in the subject being taught, mainly through the use of the tool itself. The evaluation method proposal presented in this work focuses on finding the extrinsic motivational power of a tool, which does not exclude the help of intrinsic motivation by its audience. There are several kinds of interfaces, as mono/multimodal graphic, mobile, tangible, haptic, organic, gestural, sonorous, holographic and brain-computer. Our method intends to embrace them all, once related to educational tools, bringing a set of heuristics that can be used to several interaction contexts – being more general in this side – since the focus of the tool is an 109 Chapter 7. Evaluation of Motivational Aspects

interactive product for educational purposes. Still, there are several kinds of educational applications, as tutor systems, exercises/practice systems, games, quizzes, and the method is intended to fit them all too. The heuristic evaluation is a method commonly used to evaluate usability issues of a product; it is considered an easy, fast and cheap – if compared to others (e.g. cognitive modeling methods, pluralistic inspection, focus groups) – method to be used in an evaluation process, and typically used to evaluate usability [105]. It is based on following heuristics, which are sorts of guidelines that indicate possible issues in an application from a given context or point of view. The work of Omar and Jaafar [81] presents a method to evaluate playability of games with heuristic evaluation, which shows the possibility of using heuristics for evaluation of different aspects; inspired with this possibility, the present work presents a method which uses heuristics as basis to evaluate motivation in educational tools.

7.3 Method for Evaluation of Motivational Aspects

The evaluation method is a heuristic-based technique focused on evaluation of motivational aspects in the context of educational media and technology. The Method for Evaluation of Motivational Aspects has the purpose of evaluate a technological educational tool by the user perspective of how powerful is the influence of the given tool in motivate the student-user to keep studying the subject embraced by that tool. This evaluation method adopts as a basis the ARCS model [98][100], which divides motivational aspects in four axes, i.e. attention, relevance, confidence, and satisfaction. A brief explanation [100] of each one of the axes clarifies that Attention is the aspect that brings the student to use a given learning tool, i.e., what engages curiosity, and also, variety is important is this aspect, to maintain the attention after the first impression. Relevance is to bring to the student a relation with meaning with the study, presenting a content that is important for them; this relation can be in personal or professional areas of the student’s life. Confidence is to provide clear objectives to the students, and challenges possible to be accomplished, in order to help students establish positive expectancies for success. And Satisfaction is to provide fair rewards to the student, according to their performance in learning. The four axes work together to raise the motivational power of a given course or learning tool. Our heuristic-based evaluation method is intended to fit all branches of technological educational tools, once it is an embracing method, where the heuristics will guide the evaluators to bring out questions and issues focused on each one’s context. The ones who will apply the method are guided through the heuristics for the development of a set of questions using the heuristics that fit more in the given context, and they should direct these questions to embrace specifically the tool to be evaluated. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 110

The minimal number of experts to apply the method is one, and they should be familiar to the tool that will be evaluated, as to its goals. It is desirable that the experts have knowledge about heuristic evaluation, as well as about the whole context of the tool being evaluated; this can be achieved with people that have expertise in all of these topics, or with a crew of experts in which each one is responsible for a determined expertise. The application of the method is done in the following fourteen steps grouped into three stages: i) pre-experience (accomplished by the experts); ii) during-experience (with the participation of both experts and prospective/end users); and iii) post-experience (again only the experts). The steps inside each phase (pre, during and post-experience) do not necessarily follow a mandatory sequence, once the steps are complementary. The defined sequence is to do first the pre-experience steps, then the during steps, and finally the post steps; but inside these phases, the steps’ accomplishment sequence can be adapted the way it fits better in the context, or as the experts find it better to achieve the results. Figure 1 represents all steps of this evaluation method; the steps highlighted in bold and italic are specific from the motivation evaluation, once the other steps can be found and followed in other kinds of evaluations; and next, we present an explanation of every step. Above each one of the three phases, it is indicated who participates, i.e. (E)xperts and/or (U)sers.

Fig. 19 - Steps of the method. First, there is the pre-experiment stage of the evaluation method. In the User Profile Definition step, the experts search for potential or end users of the tool, based on their knowledge about it, and make de invitations. With the profile delineated, the User Formulary Development step can be accomplished, which is the development of a formulary for the users to answer about themselves, before answering anything about the tool; this formulary is mainly intended to identify the level of the user concerning the topic to be taught by the tool, and the tool itself; together with this formulary, the consent formulary is developed, which contains explanations for the user about what information they are transferring to the experts and how this data will be used (for example, personal information, photos, audio and video recordings). The Final Interview Development step consists in developing a set o questions by the experts to be asked verbally to the user in the last step of the during-experience time; the 111 Chapter 7. Evaluation of Motivational Aspects

questions should include topics about the experience as a whole, and also complementary questions about the evaluated tool, giving the user the opportunity to express better some things they did not expose in the motivational questionnaire. In the Motivational Questionnaire Development step is the last step of the pre-experience, and the experts are responsible for developing a questionnaire guided by the heuristics (Table 16); the questions developed should follow a format, which is a question with two kinds of answer: a Likert-type scale (1-5) and a space for comments, being the former mandatory and the latter optional for the user/evaluator to fill. For each axis of ARCS model [98, 100] – i.e. attention, relevance, confidence, and satisfaction –, we recommend to develop a minimum of two questions, but this number is extensible/adapted for the needs of the context, being careful to not become the experience tiring, since the questionnaire answering is just one of the five steps that the users will participate. Table 16 shows the 31 heuristics divided in four axes, i.e. Attention (9), Relevance (6), Confidence (7), and Satisfaction (9). Table 16 - Map set of heuristics developed for use with the method. Heuristic Axis # Description A1 Graphical interface captures user’s attention A2 Design stimulates user’s inquiry

A3 User interaction maintains the interest of the user A4 Interaction between users stimulates continuous usage of the tool A5 Multimedia used maintains the interest alive A6 Tangible features draw the attention to the tool at first (A)ttention A7 Presentation of content offers interesting and innovative visual and features A8 Interaction is intuitive A9 Interface is practical, clear and objective

R1 Modes and levels embrace the various kinds of users

R2 Learning goals are according to the profile/needs of the users R3 Content is relevant to the target audience R4 Tool relates teaching with the experiences of the users (students)

R5 Interface/interaction features are familiar to the audience (R)elevance R6 The goals are stepwise and well guided in the flow of the tool C1 Learning goals are clear C2 Learning/testing levels follow the user’s development C3 Goals and levels can be configured to attend several levels reached by the user C4 Tool offers different kinds of rewards C5 Rewards do reflect the effort/ability level of the user

(C)onfidence C6 A way of learning measurement is comparing results from other users C7 A way of learning measurement is to maintain status of the user and evaluate single progress S1 The user has the possibility of test/perceive, through the tool, the user own progress S2 Social interaction function between users provides/shares rewards by success

S3 Rewards’ measurement is fair, according to the level of the user S4 Learning and progress are reinforced/highlighted when the accomplishment of specific goals S5 Tangible/gestural features are ergonomic and not tiring S6 Feedback is well presented, in a way the user can analyze the progress (S)atisfaction S7 Interaction is inclusive for several kinds of users, considering motor abilities S8 Interaction is not exhaustive after a considerable or desirable time of usage S9 Interface is not mentally exhausting

Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 112

Starting the during-experiment stage, is the User Formulary Application step, in which the consent formularies are presented and explained, together with a quick explanation about the experience as a whole, and then, if the user agrees, the user profile formulary is applied, intending to know better the profile of each user individually. The Tool/Prototype Presentation step consists of presenting the tool to be evaluated to the users, for guarantee that they all will be able to answer the questionnaire with property; the duration time of presentation is decided by the experts, according to the context, amount and complexity of functions and learning features. Once presented the tool, next is the Tool/Prototype Usage step, in which is given the user some time – determined previously by the experts, and equal to every user – to explore the tool as a whole; also, the experts can ask the user to accomplish some tasks, also defined previously. Still, in the step, experts should note user’s behavior, doubts, questioning and other interesting things that may occur during this step; this observation is very important and can even help the final interview to flow better. In the Motivational Questionnaire Application step, the heuristic-based questionnaire – developed in the first step of the method – is applied, for the users who are participating in the process; the time to answer should be determined by the experts. In the Final Interview Application step, the questions developed in step IV are the guide, but depending on observation of the experts about the behavior of the user during the usage of the tool, these questions can be adapted, or changed of order to give better flow in the conversation; the interview should be recorded – with audio and video – and also, if possible, the expert or some auxiliary person can make some annotations about the answers. The post-experiment stage begin with the Quantitative Data Analysis step, in which the experts analyze the answers of the motivational questionnaire, using the Likert-scale levels to generate a severity table about possible tool issues detected in these answers; also, flow and usage diagrams are generated, which will contribute to infer thing when crossing this data with profile and other extracted information. In the Qualitative Data Analysis step, the experts should read and listen to all the information recorded during the experiment, plus the subjective answers of the users, in order to organize the data to help the future requirements analysis, and also to contribute in the inferences of the next step. The Data Gathering step will unite qualitative and quantitative data. The experts should cross information of the generated representations with the information collected and organized in previous steps, aiming to deduce the reason of the answers – being good or bad –, compare the final results of each user based on each profile, and help to decide what decisions to make in the final steps of the method, i.e., find the positives and negatives and define which requirements should be inserted, modified and/or withdrawn. The Motivational Strengths and Weaknesses Documentation step is for the experts to use the data gathered to develop an objective list that describes the stronger and weaker points of the tool under a motivational view; this list will help guide the Requirement Analysis step. In the Requirements Analysis step the experts analyze the 113 Chapter 7. Evaluation of Motivational Aspects

initial requirements of the tool, identifying issues and gaps according to the evaluation results; and extract new requirements by analyzing suggestions and comments of the users. At the end of the process, the experts will be able to analyze – through the questionnaire answers (quantitative) and comments (qualitative), the interview and the observations during the usage – the results for their tools related to motivation. Through this data, which is focused on the users’ opinion, it is possible to conclude the initial motivational power of the given tool. Also, it is possible to do another instance of the evaluation, with the same users, to ascertain the motivational power of the tool along the time for the same person. The method can show issues that can be related to other aspects, as usability and interface, but in a way that contributes to the final motivation of the user, i.e., these issues will be analyzed under the motivational view, once they leverage the motivational power of the tool.

7.3.1 Example of Application of the Method

Following the steps of the method, as described in the previous section, we present in this subsection an exemplification of each motivational step of the three phases, since it is the main contribution of our evaluation method; we focus in explain how to accomplish these steps. It is important to contextualize the example, which was made with an educational game called Karuchā Ships Invaders [80] that aims to support basic Japanese language learning, including a basic alphabet – i.e. hiragana – and cultural words. The example described below was based on the case study that was previously conducted to evaluate the method described in this work. First is the Motivational Questionnaire Development step, and here we present four example questions, one for each motivational axis, i.e. attention, relevance, confidence, and satisfaction. To develop these questions, the expert has to have in mind the aspects of the tool that he wants to evaluate, using the general heuristic as a guide, but developing a question that describes specific points of the given tool. The questions were developed in multiple choice format, and, additionally, at the end of each question there was a field for optional comments that users could fill in if they desired, in order to explain about their answer or make suggestions. Still, it is important to explain that it not mandatory to use all of the heuristics to develop the questions, once for some contexts, one or some of them may not fit; an example, in the context of Karuchā Ships Invaders, is the S5, which embraces features the game does not offer, as tangible interaction. Table 17 shows the example questions for Karuchā Ships Invaders. Table 17 - Example questions guided by the heuristics for the example of the method. Axis/Basis Heuristic Closed Question Answers (cf. Table 1) 1) Not at all A5.1. The images used are 2) Just some elements (A)ttention A5 pleasant and in the learning 3) Pleasant, but out of context of Japanese context of Japanese? 4) In the context of Japanese, but not pleasant 5) Very pleasant and all in learning context of Japanese Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 114

1) Not at all R3.1. The content Karuchā Ships 2) No, the content is low (R)elevance R3 Invaders intends to teach fits your 3) Somewhat, it is very specific (user) needs? 4) Yes, but I’d like more content options 5) Yes, all I want to learn 1) Not at all C2.1. Your development in 2) Just in the beginning (C)onfidence C2 learning was accompanied by the 3) Just in only-kana levels levels of Karuchā Ships Invaders? 4) Yes, mostly 5) Yes, completely S4.1. Were you able to 1) No, nothing in the game highlights that reinforce/highlight what you 2) No, because it is repetitive already learned by the (S)atisfaction S4 3) Yes, but only if I focus strongly on the learning accomplishment of some task 4) Yes, naturally (some level completed or 5) Yes, every time I complete a level (or mode) repeated, for example)?

The next motivational step is the Motivational Questionnaire Application. In this step, the user was asked to answer the motivational questionnaire, explaining the multiple choice questions and that the comments’ space was for complementation or observations, but optional. Still, it was explained that during this activity, the user could ask for the experts to clarify any doubts about the questionnaire answering; it is important to highlight that the experts are present and active in all steps of the whole process. Also, at the end of the questionnaire, there was a space for more comments and suggestions for the game. The last step dedicated to motivation is the Motivational Strengths and Weaknesses Documentation, in which the experts will analyze the data gathered and extract this information. Table 18 presents some example results in the context of Karuchā Ships Invaders, also divided by ARCS axes. These results were achieved by analyzing the questionnaire answers, the comments, the observations during the usage and the users’ interview answers. They are presented in natural language and indicate issues that will guide the next step – Requirements Analysis – with suggestions of features to be treated, improved, included or removed from the tool in future iterations, to better attend the users’ educational needs. Table 18 - Examples of motivational strengths and weaknesses. Axis Comments / Suggestions / Issues 1. Some images chosen for the cultural words (bosses) are difficult to understand/read. Attention 2. The game design is very pleasant/beautiful. 3. The content to be learned fit well for most of the users’ profiles. 4. New modes were suggested: a mode in which the most missed elements fall with more Relevance frequency, for the user to practice more his weakness; a mode with possibility to customize what gyos/kanas will fall in the level. 5. At the end of the level, the missed kanas should be presented with its reading (roma-ji) beside it, instead of only the Japanese character. Confidence 6. Several users approved the stats and would like to share the results and compare it with other players’. 7. There is a lack of an elaborated scoring system, with balance, that reflects the effort and performance of the user. Satisfaction 8. The portability for other operating systems, some users commented that it would be unlikely to change their preferred operating system just to play Karuchā Ships Invaders.

115 Chapter 7. Evaluation of Motivational Aspects

7.4 Discussion

First, it is important to highlight that the method proposal was refined through the accomplishment of a case study and post analysis of the experience; the case study was accomplished with an educational game called Karuchā Ships Invaders [80], and the examples presented in the previous section are based on this experience. Initially, the method only had 7 steps, which were improved and incremented with the help of the pilot experience. The pilot also helped us to improve each step with details that were only perceived in practice, for example, the need of written consent terms, the time of presentation of the experience for the user and the details that should be given. Still, the most increased phase was the post-experience, which only contained one step, and now there are five; we were not capable to describe the data analysis in such details until we had to actually perform it. The separation of tasks for the analysis proved quite necessary, once the data collected is considerably wide and heterogeneous. The elaboration of the motivational questionnaire presented some difficulties, as the fact of how to use terms that were not too technical, for common users to be able to understand and answer the question properly. We indicate that a pilot test be accomplished with focus only on this point, before the usual pilot experience. Also, depending on the scenario, the heuristics can fit more or less, making it respectively more easy and more difficult to develop a set of questions to embrace the whole tool features; and as a result there will be bigger and minor questionnaires depending on the context and on the effort, knowledge and available time of the experts to develop them. For that, the method can extract/detect more or less issues of the tool. The method has a considerable set of heuristics intending to embrace the most cases as possible, and the ability of the experts, as the user profiles and even the end users who actually participated of the experience are determinant factors to the success of the application of the method. The evaluation method also provides considerable guidelines to the whole process, which can help inexperienced evaluators too. The definition of user profile is crucial, it has to be very well delineated, to match the interests with what the tool offers, once this match will impact on the answers given and the will of the user to cooperate properly. The success in developing a useful interview also depends on these first steps. During the experience, it is important to provide an equal environment for all the users, following always the same steps, presentation and rules for everyone. That is why is it so important to accomplish pilot tests, first with the questionnaires, to evaluate if they are clear and extract the information the experts want to; and also with the whole experience, to detect previously possible issues that may interfere and/or impair the results. The data analysis is a very challenging activity, because the data should be crossed: the results of the questionnaire, the interview answers and behavior, and also the user profile formulary answers. But this analysis can extract several aspects of the evaluated tool; the Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 116

comments made by the users may embrace issues and suggestions for interface and interaction features, technological platforms, the experience itself, among others; because these themes have a relation with the motivational impact of the tool at the end. A user may find the tool interesting, but it also may not have possibility to access it after the experiment (e.g., because of a different operating system, computer hardware, and specific devices). Finally, it is fundamental to highlight that the results obtained with this method proposal showed that it needs to be revised, with a deeper knowledge of motivation as a whole, aiming to improve it to obtain a more closed method which presents more clear motivation results at the end of its usage. This method showed to have much potential, and an expansion and/or increase, or even a more specific scope, can make it become even more useful to evaluate motivational aspects in a determined learning tool context.

7.5 Conclusion

This chapter presents and exemplifies a heuristic-based evaluation method proposal to evaluate motivational aspects of technological educational tools. This method uses a division for motivational aspects in four axes to create heuristics that guide evaluators in the development of a set of questions that are able to extract several issues about the motivational power of a given tool. We described this proposal in detail and presented some examples of how to apply it, in order to support others to use this method, although we highlight that it still needs improvement. As future work, we intend to expand this work with a deeper conceptual basis about motivation, aiming to obtain better and more accurate results with the application of the method improved. Also, we indicate a motivational evaluation using educational tools, with different themes and learning aspects.

7.5.1 Decision Making

Through the accomplishment of this step of the research, the next one is highly united to this. The decision made is to analyze the results of this experiment validation, but over the point of view of the results for Karuchā Ships Invaders game. The next step of this research is to compile all the data collected in order to improve several aspects of Karuchā, analyzing and attending, to the extent possible, to the users’ suggestions.

Chapter 8 8 Applying Evaluation of Motivational Aspects on Karuchā Ships Invaders

8.1 Introduction

CALL (Computer Assisted Language Learning) tools are increasingly present in our daily life. There are several kinds of tools that can be classified as CALL - e.g. websites, desktop software and mobile apps - and for each of these kinds, there are several teaching approaches. It is possible to find tutor systems, exercises collections, cooperative online learning, and also games. CALL games are very interesting and engaging for most users, since it blends learning with entertainment and fun. The present work focuses on CALL games and their influence on the motivation and learning of the students, including the development and expansion of evaluation methods to obtain this kind of data. The evaluation of motivational aspects of educational games is important to give the developers several clues about the end-users’ approval or disapproval for the game features, and guide the course of development. This work is placed in the context of a research project whose theme is Japanese language learning supported by computational tools. Therefore, this chapter presents new results of our research on this subject. In previous chapters, we presented the state of the art (chapter 2), and the state of the practice of web (chapter 3) and mobile (Android) (chapter 4) platforms of CALL tools for Japanese language. Also, we developed a CALL game called Karuchā Ships Invaders (chapter 5) focused on Japanese language learning, evaluating its interface aspects (chapter 6); and a method of evaluation of motivational aspects for educational media and technology (chapter 7). Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 118

Karuchā Ships Invaders [80] is a CALL game48, in a space invaders’ style, focused on Japanese language, developed by our research team in Lua programming language with MOAI49 game engine platform. With it, the student-player can learn and practice knowledge about the most basic Japanese alphabet – i.e. hiragana50 – and cultural words related to Japanese culture. The game tells a story where Japanese ships are heading to Brazil and need help to land. To help them, the player has to know the basics of the Japanese language, to pass the commands to the ships. The game offers three basic difficulty modes, i.e. easy, normal and hard, each one increasing velocity and frequency to the ships’ fall. Also, it exhibits stories to present the content to the players – this is one of the clear educational aspects of Karuchā. In this chapter, we present the application of the method developed earlier on Karuchā Ships Invaders educational game and its results regarding the motivational power of our game in making Japanese language students want to keep their learning. The experiment was accomplished with Brazilian prospective users and provided quantitative and qualitative data, whose analysis will also be detailed in the present chapter. This chapter is organized as follows: subsection 8.2 introduces the research and methodology scenario; subsection 8.3 presents the results for the evaluation; subsection 8.4 presents the discussion; and subsection 8.5 concludes the chapter.

8.2 Research and Methodology Scenario

The method is a heuristic-based method for evaluation of motivational aspects of educational media and technology. The heuristics of the method were developed based on the ARCS Model [98], i.e., divided into four axes: Attention, Relevance, Confidence and Satisfaction. ARCS Model was first developed to be used in traditional courses in Educational, but there are works that used it as a basis to accomplish evaluations with digital educational programs, as in Shellnut et al. [106], Wongwiwatthananukit [107] and Huang et al. [97]. The experiment was divided in six stages: definition, planning, data gathering, calculation, presentation and data analysis and interpretation. Follows the experiment guide for the Data Gathering stage (when the users do participate), which was divided in five steps, to detail and clarify the process: 1. There is a quick presentation of the environment, in which the user is aware where to stay during the experience, what will be the steps of the experiment, know about the cameras

48 Available for free download at http://www.karucha.pairg.dimap.ufrn.br. 49 http://www.getmoai.com. 50 Hiragana is the most basic Japanese alphabet, used for native words for which there are no kanji, or an unusual kanji, and also for grammatical particles, suffixes, verb and adjective inflections. Hiragana is also used to show the pronunciation of kanji as a reading aid, like a subtitle, what is called furigana. 119 Chapter 8. Applying Evaluation of Motivational Aspects on Karuchā Ships Invaders

and fill a Free and Clarified Consent Form, in order to allow the experimenters to collect the data. 2. The experimenter follows a presentation guide (whose accomplishment should take up to 10 minutes): a. Explain about Karuchā (download, plot, educational features and rules); b. Explain about the motivational testing (where the user’s ability or Japanese language level are not being evaluated/judged, but the main goal is to determine the motivational power of the tool); c. Present the game menu, options and keyboard shortcuts; d. Present the gameplay, playing a level as an example. 3. The user is free to explore the game for up to 40 minutes (it is also explained that if the user desires to stop using the tool, they are able to do that at any moment). In this step, the experimenter note observations about the user-tool interaction. 4. The user answers the motivational questionnaire, being aware that they can ask for help due to possible doubts. 5. The experimenter mediates the final interview with the user, based on the pre-defined questions, but being able to adapt the flow accordingly to observations about the user during the usage of the game.

For this evaluation, we needed users who were involved somehow with Japanese language and culture. We searched between students and also professors; and, since it is a game we are evaluating, we searched for the users among people who were already computer game players and people that are not (searching for a player and a non-player for each of the profiles). The user profiles we searched for to accomplish the evaluation were: i) aspirants to the study of Japanese language (no previous knowledge); ii) beginners in the study of Japanese language (some knowledge in kanas51); iii) intermediate students of Japanese language (JLPT52 N5 or some kanji53); iv) advanced students of Japanese language (JLPT N4 or N3); and v) expert students of Japanese language (JLPT N2 or N1). These profiles were chosen because we would like to observe if the game is suitable for all these levels of knowledge, and also to see whether motivation varies due to the knowledge when playing the game. We were able to accomplish the evaluation with 11 users, one at a time. The de-tailed profile of the users is described in the diagrams (Fig. 20), which detail the following information:

51 Kanas are the two basic alphabets of Japanese language, i.e. hiragana and katakana, which compose the former reading and writing studies of the language. 52 JLPT is the Japanese Language Proficiency Test, divided in five levels, in which N5 is the easiest and N1 is the most difficult level. 53 Kanji are ideogram characters of the Japanese language, more complicated than kana, and much more numerous (each kana alphabet has 48 basic symbols, whilst kanji are 2,136 of most usage (taught mandatorily), but the total amount is not exact. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 120

Japanese knowledge of the users in the four language skills – i.e. writing, reading, speaking and listening – and Japanese cultural knowledge. From all the 11 users, the age varies from 16 to 23 years old; 73% are male and 23% are female; and 91% declared themselves as computer game players, against 9% non-players.

Fig. 20 - Diagrams that detail user characteristics.

8.3 Results

8.3.1 Quantitative Results

The quantitative results will be presented in tables and figures with diagrams, divided into Log Flow and Results of Severity Analysis of the Motivational Questionnaire.

8.3.1.1 Log Flow

The following figures represent the flow of the users through the game, recorded in the log file. Fig. 21 shows the maximum level each user achieved, in each one of the three most basic difficulty modes, i.e. easy, medium and hard54. Fig. 23 shows a flow of a specific user (U3) through levels of the hard mode, showing how many times he played, failed and won each level and the maximum level reached. Fig. 22 is a counter of how many times each user used the context-help55 in each level.

54 Each difficulty mode has 30 levels, in which every three levels there are bosses mixed with hiragana, and the levels 15 and 30 are only-boss levels. 55 Context-help is a feature with educational purposes, in which the student-player can consult, during the level, the reading of the kanas or bosses. It is available in easy mode for all the kanas and bosses presented so far, and in medium mode, only for the kanas and bosses of the present level. 121 Chapter 8. Applying Evaluation of Motivational Aspects on Karuchā Ships Invaders

An interesting thing to analyze about Fig. 21 is that few users played all the difficulty modes, and crossing data with the profile, the users who played in hard mode were the ones with more knowledge about Japanese language. Another relevant observation is that a user (U9) with no knowledge about the idiom was able to achieve a high level (level 19) in easy mode, and crossing data with Fig. 22, it is possible to infer that the context-help feature was very important for this accomplishment. U8 and U10 also fit in this description. The U1, instead, took another approach: crossing the diagram with observation data, it was possible to see that this user was trying to learn the symbols through several repetitions, and that is why, even taking all the given time, he only reached level 3. Fig. 23 shows a representation of a game flow through hard mode of a specific user (U3). It is interesting to notice that the levels he failed the most had bosses (level 9 and 12), but the previous boss levels (3 and 6) were won in first try. This shows that the game difficulty is increasing with the levels. Also, it is perceived that the level he was not able to pass was an only- boss level (level 15). Even if the user is an intermediate Japanese student, the game was able to provide a challenge, showing that in fact it suits since beginners until advanced Japanese students. About Fig. 22, the main inference is the relevance of the context-help feature. Only three of the users did not access the feature, being two of them intermediate/advanced Japanese students. The number of times this feature was used is noticeable, in a period of 40 minutes, there were users able to access it for about 150 times. Except for the ones who did not used context-help, the minimum of times a user accessed it during the experiment was about 30 times. This feature was not originally in the initial concept of Karuchā, but the importance of its addition to the experiment is highly proved by this diagram, and also by the logs, observations and users’ answers.

Fig. 21 - Maximum level the users achieved. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 122

Fig. 22 - Counter of use of the context-help in each level.

Fig. 23 - A flow through levels of hard mode, from user “U3”.

8.3.1.2 Results of Severity Analysis

Fig. 24 is a table containing the evaluation of severity analysis for the motivational questionnaire, where there is represented an average of each question’s response including all the users’ responses. The first group is for Attention, second for Relevance, third from Confidence and fourth from Satisfaction. It is noticed that the lower scores were related to Confidence and Satisfaction, and the analysis of what may have caused it is presented on qualitative results. Still, crossing profile data with the answers it is perceived that lower scores were given by advanced Japanese students or regular computer game players, which are more rigorous users in this 123 Chapter 8. Applying Evaluation of Motivational Aspects on Karuchā Ships Invaders

context. Their answers were very important and should be well analyzed in order to improve the motivational aspects of Karuchā educational game. Question U1 U2 U3 U4 U5 U6 U7 U8 U9 U10 U11 Min Max Mode Average 1 5 5 4 4 5 4 5 4 5 4 5 4 5 5 4,55 2 5 5 3 4 5 3 4 4 4 3 4 3 5 4 4,00 3 5 4 4 4 4 5 4 5 4 4 4 4 5 4 4,27 4 4 5 5 5 5 5 5 5 5 5 5 4 5 5 4,91 5 4 3 5 3 4 5 3 3 5 5 4 3 5 3 4,00 6 4 5 3 3 5 5 3 3 3 3 3 3 5 3 3,64 7 4 4 3 4 3 4 4 5 5 3 3 3 5 4 3,82 8 5 4 3 5 5 5 2 3 5 5 3 2 5 5 4,09 9 4 4 2 4 5 5 2 5 5 5 4 2 5 5 4,09 10 5 5 2 4 4 5 3 5 5 5 4 2 5 5 4,27 11 5 4 4 5 4 3 3 3 4 4 2 2 5 4 3,73 12 4 4 4 4 4 5 4 4 4 4 2 2 5 4 3,91 13 5 4 5 5 5 5 5 4 4 4 4 4 5 5 4,55 14 5 5 5 4 5 5 3 3 5 5 5 3 5 5 4,55 15 3 5 1 3 4 3 4 3 2 3 2 1 5 3 3,00 16 5 5 1 3 5 5 3 3 3 3 5 1 5 5 3,73 17 4 3 4 3 4 5 4 3 4 3 5 3 5 4 3,82 18 4 4 1 4 5 5 3 3 4 4 4 1 5 4 3,73 Fig. 24 - Severity table of the motivational questionnaire’s answers.

8.3.2 Qualitative Results

The qualitative results of this experiment were extracted from observations during the usage step (step 3), comments written spontaneously by the users in the motivational questionnaire (step 4) and interview answers (step 5). Through this data, it was possible to confirm some issues the developers were already aware because of the interface evaluation, that were not fixed until the experiment. For example, issues with size and choice of some bosses’ pictures, commented by some of the participants. Also, some animation effects were discovered as giving the wrong message, such as the abduction animation, that could pass the idea of destruction of the ships, instead of the intended help to land. A positive fact detected about the design is that several of the users demonstrated more interest for the game after seeing its first graphics, as the menu screen and a level screen. Other issue detected through users’ suggestions was in the context-help feature; this feature’s importance and relevance was already confirmed in the quantitative data analysis, but the users gave ideas to improve it. One of the ideas is to offer help of all levels until the current level, being available to navigate through levels’ new kanas. Also, it was suggested a boss context- help, following the same format of kana context-help. Issues related to statistics presented at the end of each level were also detected. It was suggested that the stats presented not only the missed kanas, but also their reading in roma-ji, i.e., the roman letters. Some users also commented they would like to compare their results with other student- players, in order to get extra motivation to improve their skills in the game. Another interesting feature suggested was to the game give recommendations to the player, depending on how the level was finished. For example, if the player won the level, but took too many hits, the game Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 124

could give a message like “You may need to repeat this level to improve your learning”. Other suggestions embraced the development of an elaborated scoring system, to be used for comparison/competition between users; the development of online server and features, such as achievements, timeline stats, social share of results; and also, portability to other OS, such as Linux and Mac. All these punctual repairs contribute strongly to improve the motivational power of Karuchā, since every comment that is analyzed and treated properly makes users more comfortable in using the game, and the flow becomes more pleasant. The following figures show some user reactions during the accomplishment of the evaluation. While playing the game, interaction with the game was registered (Fig. 25), and also users’ facial expressions (Fig. 26); moments of the motivational questionnaire and interview answers are showed in Fig. 27. It is noticed that various users had good moments during the usage of the tool, which is mainly perceived in Figs. 25 and 26, through visual expressions56 of joy. Also, there were moments of confusion, doubts and frustration, which were caused sometimes for interface issues, and sometimes for the game itself, where the player can win or lose.

Fig. 25 - Users playing Karuchā Ships Invaders.

Fig. 26 - Users’ reactions to the game.

56 All the users signed a Free and Clarified Consent formulary, allowing the recording and use of their images for sake of experiment’s results. 125 Chapter 8. Applying Evaluation of Motivational Aspects on Karuchā Ships Invaders

Fig. 27 - Users answering motivational questionnaire and interview.

8.3.3 New Features

Following, we present a list of features classified by the motivational axes, and extracted from the users’ answers - optional comments and interview - in the experiment. There were 20 new features identified in this activities with student-players: Attention (FA1..FA8), Relevance (FR1..FR5), Confidence (FC1..FC3), and Satisfaction (FS1..FS4).  Attention: FA1. Improvement of bosses’ images; FA2. Improvement of ship abduction effect; FA3. Highlight the clickable areas; FA4. Implementation of other modes, such as survival; FA5. New modes/levels with different time limit; FA6. Keep music during context-help usage; FA7. Option buttons available in the levels; FA8. More variety of sound effects.  Relevance: FR1. Context help of previous levels on easy mode; FR2. Context help of bosses; FR3. More intuitive keyboard shortcuts; FR4. Make the most missed elements to fall with more frequency (a new mode); FR5. Implementation of a customizable mode (choose what symbols to practice).  Confidence: FC1. Show roma-ji for missed kanas at the end of the level; FC2. Comparison of results with other players; FC3. The game give recommendations, such as “you may like/need to repeat this level” if the player misses many ships.  Satisfaction: FS1. Implementation of a scoring system, different from number of hits; FS2. Portability to other OS; FS3. Implementation of competition/interactive mode; FS4. Development of online features (such as timeline diagrams, general stats, and achievements). Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 126

8.4 Discussion

The application of method to evaluate motivational aspects of the educational game Karuchā Ships Invaders brought to light several results from the data analysis. Analyzing the method itself, there are various questions that can be raised and discusses, such as regarding the application possibilities. In the present work, method was applied in person, and there are several characteristics of the method that can only be achieved this way, such as: usage live observations; video recording of interview; the interview itself; explanation of users’ doubts by the experimenters/developers. On the other hand, it is possible to do some adjustment in the method, in order to accomplish a future evaluation with the new online features of the game. For example, the questionnaires can be available online for the users to answer; the video for recording facial expressions can be recorded by users’ webcams, if they agree to; and doubts can be solved online with FAQs or members of the evaluation team that are available to do so. Still, it is possible to consider using another evaluation method, but adapting the method to sundry situations may be important and make more sense, once it is a way to extract similar data from the in person experiment. Though, this adaptation demands more research and study. About the game itself, there are several topics to discuss. The evaluation raised many issues that turned into new features to be implemented; which demand much work, new implementation and redesign included. The Attention axis, despite of being the one with the higher scores in the Likert-scale, was also the one who received more comments and suggestions. Several interface issues were pointed, such as the fact that some boss images contain very small or confusing representations; the ship abduction effect can make the player think he is destroying the ship instead of helping it; the clickable areas of the game are not well highlighted/indicated. These are features that will be redesigned in a next step of the research, in which the game will be completely rethought in terms of design and implementation; this is necessary to maintain the consistence of the game as a whole, once it intends to offer an immersive environment to Japanese language students, that would not be complete with a very well designed interface. Other new features are already solved, as keeping the music playing during context-help, more option buttons during the levels, and the new modes and levels are in design process. The Relevance axis raised considerable questions too, as the implementation of context- help for bosses, and the possibility to navigate through the helps of previous levels. These features are already implemented, as well as another help, with less hints, for medium difficulty mode, in order to reinforce the learning experience. The new modes proposed for this axis’ analysis are also in design process, as the modification of some keyboard shortcuts that the users pointed as non-intuitive. 127 Chapter 8. Applying Evaluation of Motivational Aspects on Karuchā Ships Invaders

The axis of Confidence brought few, but important new features, such as the already implemented exhibition of roma-ji for missed kanas at the end of each level. Another interesting suggestion was the idea of the game to give recommendations to the student-player, this can also reinforce the learning nature of the game. Some users also commented the desire of having their results shared and compared with other players, and this is interesting because in the interview form, a question was directed to the possibility of having an online server with some social features, though this was not implemented at all, but it was already an idea discussed by the Karuchā’s development team. The Satisfaction axis is the one whose new features are all left to implement; this axis raised more complex questions, as the development of an elaborated scoring system, the development of the game to another OS and platforms (the latter requires several thinking, once it can embrace other devices with different input characteristics), and the generation and presentation in real time of all the data, similar to what were collected in the experiment; but in the online feature, each player should have an account that saves, processes and presents all the data in graphics, diagrams and with customizable options, in addition to the option to share it to other social networks. This axis showed that there is a series of features that may seem basic and simple, but that demand complex thinking to maintain everything working and making sense, mainly in the context of an educational game, to continue offering a good balance between learning and fun, keeping also the learning context presented in the interface elements and interaction.

8.5 Conclusion

This chapter presented the application of the method for evaluation of motivational aspects of educational games, describing all the steps for one specific evaluation scenario, including results. The method was applied in Karuchā Ships Invaders educational game and the results showed that the game is a powerful influence to Japanese language students, since the beginners and extending to advanced students. The evaluation also brought to light several issues that were turned into new features to be included in the game. It is noteworthy that several of these new features are already implemented after the accomplishment of the experiment. As future work, we indicate the accomplishment of a study to evaluate the possibility of using the method away (not in person), to achieve new data from the players of the online server, and also to verify the possibility of the method to embrace evaluation of online educational games. Also, for Karuchā, we indicate the implementation of the new features that are still open, and to introduce a katakana mode in the game. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 128

8.5.1 Decision Making

After all the research done, the main decision made to be accomplished, even out of the present work/text, is the improvement of Karuchā Ships Invaders, mainly, by the availability of an online server, in which users can create accounts and have several motivational and educational resources to explore the game; these features are already being developed in beta version. Also, together with it, is the implementation of the new difficulty modes proposed, which are already being designed to fit the game. From now on, Karuchā Ships Invaders tends to grow and be improved even more, opening opportunities even to new research themes in this context.

Capítulo 9 9 Conclusões

Neste capítulo, apresentamos um resumo de cada capítulo anterior aqui apresentado, apontando sua importância e relevância para a pesquisa como um todo, como também os resultados alcançados com cada um deles e os objetivos específicos que cada um foi responsável por realizar.

9.1 Trabalho Realizado

Através do levantamento do Estado da Arte de ferramentas computacionais para auxílio ao ensino de japonês, foi constatado que apesar de existirem diversos aplicativos, hardwares e softwares, estes não estão ao alcance do grande público e não são efetivamente utilizados pelos estudantes do idioma. Já no levantamento do Estado da Prática das ferramentas web, foi possível observar os tipos de aplicações que de fato são utilizadas pelos estudantes e determinar seus pontos fortes e fracos. Foi observado sobre estes aplicativos, como um ponto fraco, a dificuldade de utilizá-los em dispositivos móveis, e sendo esta uma grande tendência atual, este ponto acaba por tornar-se de considerável importância ao se pensar no aprendizado de língua japonesa por esse tipo de meio. A partir disso, foi feito um levantamento de aplicativos para dispositivos móveis focado na plataforma Android, e esse estudo ofereceu novas perspectivas com relação às vantagens e desvantagens desse tipo de aplicação e seus contextos de uso. Os três levantamentos citados foram de vital importância para embasar a continuidade do estudo, descrita a seguir. Após realizados os levantamentos de inúmeras ferramentas CALL para o contexto de japonês, foi desenvolvido um jogo – Karuchā Ships Invaders – que traz a proposta de ensinar conceitos básicos da língua e cultura do Japão, com foco no – mas não limitado ao – público brasileiro de estudantes de japonês. A atual pesquisa realizou o desenvolvimento completo do Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais 130

Game Design, bem como avaliações de aspectos de interface e interação; e de aspectos motivacionais do jogo. As avaliações realizadas com experts e usuários em perspectiva mostraram questões a serem melhoradas na ferramenta, bem como nos métodos de avaliação, uma vez que o presente trabalho também propõe métodos de avaliação baseados em heurísticas.

9.2 Contribuições

 Capítulos 2, 3 e 4: Estes capítulos foram de vital importância para a análise das ferramentas já existentes no contexto de nossa pesquisa, bem como para avaliar pontos fortes e fracos das ferramentas atuais, ajudando a alcançar nosso primeiro e segundo objetivos específicos (“Identificação dos principais pontos fortes e fracos dos jogos atuais” e “Análise de tecnologias/plataformas”). Publicações: Marciano, J.N., Miranda, L.C. and Miranda, E.E.C. Aprendizagem da Língua Japonesa Apoiada por Ferramentas Computacionais: Estado da Arte e Desafios para a Comunidade Latino-Americana. In Proc. CLEI 2012, © IEEE (2012), 1-10. Marciano, J.N., Miranda, L.C. and Miranda, E.E.C. Japanese CALL web tools: Identifying and evaluating their multimedia behaviour in android OS. In Proc. EdMedia 2013, AACE (2013), 1087-1096. Marciano, J.N., Miranda, L.C., Miranda, E.E.C. and Pereira, R. Android MALL apps for Japanese language: identifying and eliciting interface/interaction requirements through the semiotic framework. In Proc. CLEI 2013, © IEEE (2013), 220-230.  Capítulo 5: Este capítulo foi responsável por alcançar o terceiro, quarto e quinto objetivos específicos de nossa pesquisa (“Estudo de elementos de jogos educacionais”, “Desenvolvimento do game design” e “Implementação”); nele está contido todo o processo de concepção e desenvolvimento da primeira versão do jogo Karuchā Ships Invaders, descrevendo-o em detalhes (regras, jogabilidade, prêmios, proposta de ensino). Publicações: Marciano, J.N., Ferreira, A.L.S., Correia, A.C.C., Miranda, L.C., Miranda, E.E.C. . Karuchā Ships Invaders: Cultural Issues on the Design/Development of a Japanese CALL Game made by/to Brazilians. In: XII Simpósio Brasileiro de Jogos e Entretenimento Digital (SBGames), 2013, São Paulo, SP. Anais do XII Simpósio Brasileiro de Jogos e Entretenimento Digital. Porto Alegre, RS. SBC (2013), 172-180.  Capítulos 6: Este capítulo cumpre o sexto objetivo específico da pesquisa (“Avaliação de interface e interação”). Realizamos o estudo de diversos métodos de avaliação de interface, optando pela avaliação heurística e adaptando um método já existente para realizar a avaliação de Karuchā Ships Invaders. Neste capítulo, também apresentamos os resultados extraídos da avaliação, apresentando melhorias técnicas a serem realizadas no jogo. 131 Capítulo 9.Conlusões

 Capítulos 7 e 8: A união destes dois capítulos cumpre o sexto objetivo específico da pesquisa (“Testes com usuário (motivação)”). No capítulo 7 temos o estudo do método de avaliação, que se desenvolve em uma nova proposta de como avaliar aspectos motivacionais no contexto de jogos educacionais, com foco do desenvolvimento e validação do método. Já o capítulo 8 traz os resultados do método para o jogo Karuchā Ships Invaders, com foco nos pontos fortes do jogo, nos novos requisitos (features) que surgiram devido aos comentários dos usuários e nas melhorias necessárias a serem feitas no jogo. Publicações: Marciano, J.N., Miranda, L.C. and Miranda, E.E.C.: Evaluation of motivational aspects in educational media and technology. In: 26th World Conference on Educational Media and Technology (2014), pp. 780–787, © AACE 2014.

9.3 Perspectivas e Trabalhos Futuros

A realização deste trabalho, com suas contribuições, dá-nos a oportunidade de seguir pesquisando este campo da ciência que são as ferramentas computacionais com foco educacional. Foi possível cumprir diversas metas, e ao longo desse estudo, é perceptível o quanto a área oferece em termos de oportunidade de pesquisa. As contribuições aqui geradas despertam novas questões de pesquisa que podem ser estudadas a fundo, gerando diversos outros trabalhos. Como trabalhos futuros, indicamos dar continuidade à evolução do Karuchā Ships Invaders, com a intenção de prover, para um público ainda maior, uma ferramenta de apoio ao aprendizado do idioma japonês. Também indicamos o estudo e desenvolvimento de outros tipos de ferramentas, como um sistema de tutoria, e que possam abranger outras habilidades linguísticas, como também outros aspectos da língua japonesa.

Referências bibliográficas

1. Mi, L., Ren, F. and Araki, H. Differences of Japanese kanji and kana during the memory processing. In Proceedings of the 7th International Conference on Natural Language Processing and Knowledge Engineering (NLP-KE’11), IEEE, pp. 205–208. 2011. doi: 10.1109/NLPKE.2011.6138195. 2. Mcloughlin, C., Hutchinson, H. and Koplin, M. Different media for language learning: does technology add quality?. In Proceedings of the International Conference on Computers in Education, IEEE, pp. 681–684. 2002. doi: 10.1109/CIE.2002.1186042. 3. Joseph S. and Uther, M. Mobile language learning with multimedia and multi-modal interfaces. In Proceedings of the 4th IEEE International Workshop on Wireless, Mobile and Ubiquitous Technology in Education (WMUTE’06), IEEE, pp. 124–128. 2006. doi: 10.1109/WMTE.2006.261359. 4. Kremenska, A. Technology enhanced language learning: student motivation in computer assisted language learning. In Proceedings of the International Conference on Computer Systems and Technologies (CompSysTech’07), ACM, pp. 1–6. 2007. doi: 10.1145/1330598.1330690. 5. Han, L., Liu, D. and Liu, Y. Implementation of CALL-based online Japanese learning system. In Proceedings of the International Conference on Internet of Things and 4th International Conference on Cyber, Physical and Social Computing (iThings/CPSCom’11), IEEE, pp. 585–588. 2011. doi: 10.1109/iThings/CPSCom.2011.77. 6. Tiwari, S., Khandelwal, S. and Roy, S.S. E-learning tool for Japanese language learning through English, Hindi and Tamil: A computer assisted language learning (CALL) based approach. In Proceedings of the 3th International Conference on Advanced Computing (ICoAC’11), IEEE, pp. 52–55. 2011. doi: 10.1109/ICoAC.2011.6165218. 7. Kato, Y. and Okamoto, T. Different evaluations of e-learning for Japanese academic reading between foreign students and Japanese language teachers. In Proceedings of the 7th IEEE International Conference on Advanced Learning Technologies (ICALT’07), IEEE, pp. 403–407. 2007. doi: 10.1109/ICALT.2007.123. 8. Lee, K.K. and Chong, M.P. Blended learning: a case study for Japanese language studies. In Proceedings of the 1st International Conference on Hybrid Learning and Education (ICHL’08), Springer, pp. 450–462. 2008. doi: 10.1007/978-3-540-85170-7_40. 9. Yusuke, K., Ruifan, and L. Ren, F. Designing a Japanese idiom education support system for overseas’ students. In Proceedings of the 5th International Conference on Natural Language Processing 133

and Knowledge Engineering (NLP-KE’09), IEEE, pp. 1–8. 2009. doi: 10.1109/NLPKE.2009.5313789. 10. Ochi, Y., Yano, Y., Hayashi, T. and Wakita, R. JUPITER: A kanji learning environment focusing on a learner’s browsing. In Proceedings of the 3rd Asia Pacific Computer Human Interaction (APCHI’98), IEEE, pp. 446–451. 1998. doi: 10.1109/APCHI.1998.704485. 11. Yusuf, M., Asaga, and C. Watanabe, C. Onomatopeta!: Developing a Japanese onomatopoeia learning- support system utilizing native speakers cooperation. In Proceedings of the IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology (WI- IAT’08), IEEE, pp. 173–177. 2008. doi: 10.1109/WIIAT.2008.300. 12. Liu, Y., Yin, C. and Ogata, H. Supporting Q&A in a web-based Japanese language learning environment. In Proceedings of the WASE International Conference on Information Engineering (ICIE’09), IEEE, pp. 422–425. 2009. doi: 10.1109/ICIE.2009.29. 13. Yoshida, M., Ishikawa, M. and Kaneko, K. An educational system for learning Japanese active proverbs. In Proceedings of the 32nd International Conference on Information Technology Interfaces (ITI), IEEE, pp. 409–414. 2010. 14. Yin, C., Ogata, and H. Yano, Y. Ubiquitous-learning system for the japanese polite expressions. In Proceedings of the IEEE International Workshop on Wireless and Mobile Technologies in Education (WMTE’05), IEEE, pp. 28–30. 2005. doi: 10.1109/WMTE.2005.68. 15. Lin, N., Kajita, S. and Mase, K. A multi-modal mobile device for learning Japanese kanji characters through mnemonic stories. In Proceedings of the 9th International Conference on Multimodal Interfaces (ICMI’07), ACM, pp. 335–338. 2007. doi: 10.1145/1322192.1322250. 16. Hou, B., Ogata, H., Miyata, M., Li, M. and Yano, Y. Development of web-based Japanese mimicry and onomatopoeia learning assistant system with sensor network. In Proceedings of the 6th IEEE International Conference on Wireless, Mobile and Ubiquitous Technologies in Education (WMUTE’10), IEEE, pp. 117–121. 2010. doi: 10.1109/WMUTE.2010.24. 17. Kanev, K., Oido, I., Yoshioka, R. and Mirenkov, N. Employment of 3D printing for enhanced kanji learning. In Proceedings of the Joint International Conference on Human-Centered Computer Environments (HCCE’12), ACM, pp. 165–170. 2012. doi: 10.1145/2160749.2160784. 18. Pershin, A., Fujitani, S. and Akahori, K. Evaluation of Japanese-English CALL tool supporting contextual language acquisitions. In Proceedings of the International Conference on Computers in Education (ICCE’02), IEEE, pp. 1289–1293. 2002. doi: 10.1109/CIE.2002.1186215. 19. Namatame, M., Harada, Y., Kusunoki, F. and Terano, T. Letting every pupil learn Japanese hand alphabets with VISUAL interfaces. In Proceedings of the Extended Abstracts on Human Factors in Computing Systems (CHI EA’04), ACM, pp. 1363–1366. 2004. doi: 10.1145/985921.986065. 20. Lam, J., Cheung, K.S., Ng, J. Yau, J., Seto, W. and Im, T. Students’ needs of e-courses as complement to traditional learning: A Japanese studies e-course case. In Proceedings of the IEEE International 134

Symposium on IT in Medicine and Education (ITME’08), IEEE, pp. 876–880. 2008. doi: 10.1109/ITME.2008.4743992. 21. Zhenlei, X., Sano, T., Yamamoto, H. and Kikuchi, A. Japanese accent training system for foreign students. In Proceedings of the IEEE International Conference on Computational Cybernetics (ICCC’08), IEEE, pp. 241–246. 2008. doi: 10.1109/IMTC.2009.5168599. 22. Miyazaki A. and Tomimatsu, K. Onomato planets: Physical computing of Japanese onomatopoeia. In Proceedings of the 3rd International Conference on Tangible and Embedded Interaction (TEI’09), ACM, pp. 301–304 2009. doi: 10.1145/1517664.1517726. 23. Nishino, H., Murayama, K., Kagawa, T. and Utsumiya, K. A Japanese calligraphy trainer based on skill acquisition through haptization. In Proceedings of the 24th IEEE International Conference on Advanced Information Networking and Applications (AINA’10), IEEE, pp. 1225–1232. 2010. doi: 10.1109/AINA.2010.112. 24. Sa, L., Bin, W., Yamashita, N., Hayashi, T. and Yamasaki, T. Intelligent Japanese learning system focusing on Chinese students’ weak points of katakana word listening. In Proceedings of the IEEE International Conference on Intelligent Computing and Intelligent Systems (ICIS’10), IEEE, pp. 215–219, 2010. doi: 10.1109/ICICISYS.2010.5658765. 25. Ichinohe A. and Suzuki, K. Expansion of the i-mode drill ‘the world of kanji’ with review functions for m-learning. In Proceedings of the International Conference on Computers in Education (ICCE’02), IEEE, pp. 1472–1473. 2002. doi: 10.1109/CIE.2002.1186302. 26. Mizumachi, I., Tawata, S. and Yamanaka, E. Developing Japanese CALL courseware for listening comprehension practice. In Proceedings of the International Conference on Computers in Education (ICCE’02), IEEE, pp. 1344–1345. 2002. doi: 10.1109/CIE.2002.1186242. 27. Wagner, D. and Barakonyi, I. Augmented reality kanji learning. In Proceedings of the 2nd IEEE/ACM International Symposium on Mixed and Augmented Reality (ISMAR’03), IEEE, pp. 335–336. 2003. doi: 10.1109/ISMAR.2003.1240747. 28. Stubbs, K. Kana no senshi (kana warrior): a new interface for learning Japanese characters. In Proceedings of the Extended Abstracts on Human Factors in Computing Systems (CHI EA’03), ACM, pp. 894–895, 2003. doi: 10.1145/765891.766054. 29. Solis, J., Avizzano, C.A., and Bergamasco, M. Teaching to write Japanese characters using a haptic interface. In Proceedings of the 10th Symposium on Haptic Interfaces for Virtual Environment and Teleoperator Systems (HAPTICS’02), IEEE, pp. 255–262, 2002. doi: 10.1109/HAPTIC.2002.998966. 30. Breen, J. JMdict: A Japanese-multilingual dictionary. In Proceedings of the Workshop on Multilingual Linguistic Ressources (MLR’04), ACM, pp. 71–79. 2004. doi: 10.3115/1118808.1118813. 31. Maciejewski, A.A. The student-computer interface on an intelligent tutoring system for Japanese language instruction. in Proceedings of the IEEE International Conference on Systems, Man and Cybernetics (ICSMC’89), IEEE, pp. 311–312. 1989. doi: 10.1109/ICSMC.1989.71305. 135

32. Fluckiger, F. Understanding Networked Multimedia: Applications and Technology. Prentice Hall International Ltd., Hertfordshire, UK. 1995. 33. Landreth, D. A Guide to Japanese Lesson Study. In Proceedings of the 22th World Conference on Educational Multimedia, Hypermedia and Telecommunications, AACE, pp. 2615. 2010. 34. Librenjak, S., Vuckovic, K. and Dovedan, Z. Multimedia Assisted Learning of Japanese Kanji Characters. Proceedings of the 35th International Convention. pp. 1284-1289. 2012. 35. Marciano, J.N., Miranda, L.C. and Miranda, E.E.C. Japanese Language Learning Supported by Computational Tools: State of the Art and Challenges for the Latin America Community (in Portuguese). In 38rd Latin American Informatics Conference (CLEI), 2012, Medellin, Colombia. Proceedings of the 38rd Latin American Informatics Conference. © IEEE, p. 1-10. 2012. 36. Matsushita, K., Watanuki, T., Gotoh, T., Yasuoka, H., Mackin, K.J. and Nunohiro, E. Development of Teaching Materials Using 3D CG Animation for Learning Hieroglyphic Kanji. In Proceedings of the 21th World Conference on Educational Multimedia, Hypermedia and Telecommunications, AACE, pp. 3974-3977. 2009. 37. Android, “User Interface Guidelines for App Designers,” Internet: http://developer.android.com/guide/practices/ui_guidelines/ [Feb. 12, 2013]. 38. Anki, “Anki,” Internet: http://www.ankisrs.net [Feb. 12, 2013]. 39. Baranauskas, M.C.C. and Bonacin, R. Design – indicating through signs. in Design Issues, vol. 24, n. 3, MIT Press, pp. 30–45. 2008. doi: 10.1162/desi.2008.24.3.30. 40. Edge, D., Fitchett, S., Whitney, M. and Landay, J. MemReflex: adaptive flashcards for mobile microlearning. In Proceedings of the 14th International Conference on Human-Computer Interaction with Mobile Devices and Services (MobileHCI’12), ACM, pp. 431–440. 2012. doi: 10.1145/2371574.2371641. 41. Liu, K. Semiotics in Information Systems Engineering. Cambridge University Press, 1st edition. 2000. 42. Magal-Royo, T., Gimenez-Lopez, J.L., Pairy, B., Garcia-Laborda, J., and Gonzalez-del Rio, J. Multimodal application for foreign language teaching. In Proceedings of the 14th International Conference on Interactive Collaborative Learning (ICL’11), IEEE, pp. 145–148. 2011. doi: 10.1109/ICL.2011.6059564. 43. Marciano, J.N., Miranda, L.C., and Miranda, E.E.C. Japanese CALL web tools: identifying and evaluating their multimedia behaviour in android OS. In Proceedings of the 25th World Conference on Educational Media and Technology (EdMedia’13), © AACE, 2013. 44. Miranda, L.C. Hornung, H.H. and Baranauskas, M.C.C. Adjustable interactive rings for iDTV. In IEEE Transactions on Consumer Electronics, vol. 56, n. 3, IEEE, pp. 1988–1996. 2010. doi: 10.1109/TCE.2010.5606356. 45. Miranda, L.C., Piccolo, L.S.G. and Baranauskas, M.C.C. Artefatos físicos de interação com a TVDI: desafios e diretrizes para o cenário brasileiro. In VIII Simpósio Brasileiro de Fatores Humanos em Sistemas Computacionais (IHC’08), ACM, pp. 60–69. 2008. 136

46. Neris, V.P.A., Almeida, L.D.A., Miranda, L.C., Hayashi, E.C.S., and Baranauskas, M.C.C.. Towards a socially-constructed meaning for inclusive social network systems. In Proceedings of the 11th International Conference on Informatics and Semiotics in Organisations (ICISO’09), Aussino Academic Publishing House, pp. 247–254. 2009. 47. Pereira R. and Baranauskas, M.C.C. Valuation framing for social software: a culturally aware artifact. In Proceedings of the 13th International Conference on Enterprise Information Systems (ICEIS’11), SciTePress, pp. 235–244. 2011. 48. Piccolo, L.S.G., Melo, A.M. and Baranauskas, M.C.C. Accessibility and interactive TV: design recommendations for the brazilian scenario. In Proceedings of the 11th IFIP TC13 International Conference in Human Computer Interaction (INTERACT’07), Springer, pp. 361–374, 2007. doi: 10.1007/978-3-540-74796-3_34. 49. Santana, V.F., Solarte, D.S.M., Neris, V.P.A., Miranda, L.C., and Baranuskas M.C.C. Redes sociais online: desafios e possibilidades para o contexto brasileiro. In Anais do XXIX Congresso da Sociedade Brasileira de Computação (CSBC’09), SBC, pp. 339–353. 2009. 50. Tian, F., Lv, F., Wang, J., Wang, H., Luo, W., Kam, M., Setlur, V., Dai, G. and Canny, J. Let’s play Chinese characters: mobile learning approaches via culturally inspired group games. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI’10), ACM, pp. 1603–1612. 2010. doi: 10.1145/1753326.1753565. 51. Verma, P. Mawl: mobile assisted word-learning. In Proceedings of the 13th International Conference on Human Computer Interaction with Mobile Devices and Services (MobileHCI’11), ACM, pp. 699–701. 2011. doi: 10.1145/2037373.2037489. 52. Wong, L.-H., Chin, C.-K., Tan, C.-L., Liu, M. and Gong, C. Students’ meaning making in a mobile assisted Chinese idiom learning environment. In Proceedings of the 9th International Conference of the Learning Sciences (ICLS’10), ACM, pp. 349–356. 2010. 53. Zervas, P. and Sampson, D.G. Enhancing educational metadata with mobile assisted language learning information. In Proceedings of the 10th IEEE International Conference on Advanced Learning Technologies (ICALT’10), IEEE, pp. 269–273. 2010. doi: 10.1109/ICALT.2010.79. 54. Aparicio, A.F., Vela, F.L.G., Sánchez, J.L.G., and Montes, J.L.I. Analysis and application of gamification. In Proceedings of the 13th International Conference on Interacción Persona- Ordenador (INTERACCION’12), ACM, pp. 1–2. 2012. doi: 10.1145/2379636.2379653. 55. Kana Attack, “Kana Attack,” Internet: http://itunes.apple.com/br/app/kana- attack/id451690838/ [Sep. 02, 2013]. 56. Kana Attacks, “Kana Attacks,” Internet: http://www.sourceforge.net/projects/kanaattacks/ [Sep. 02, 2013]. 57. Kana Drop, “Kana Drop,” Internet: http://itunes.apple.com/br/app/japanese-characters- kana-drop/id556460406/ [Sep. 02, 2013]. 58. Kana Drop, “Kana Drop,” Internet: http://www.gamejolt.com/games/arcade/kana- drop/12325/ [Sep. 02, 2013]. 137

59. Kana Invaders, “Kana Invaders,” Internet: http://www.easyjapanese.org/invaders.html [Sep. 02, 2013]. 60. Kana Invaders, “Kana Invaders,” Internet: http://www.learnjapanesepod.com/kana- invaders/ [Sep. 02, 2013]. 61. Kana Invaders, “Kana Invaders,” Internet: http://www.sourceforge.net/projects/kanainvaders/ [Sep. 02, 2013]. 62. Marciano, J.N., Miranda, L.C., Miranda, E.E.C., and Pereira, R. Android MALL apps for japanese language: identifying and eliciting interface/interaction requirements through the semiotic framework. In Proceedings of the XXXIX Latin America Conference on Informatics (CLEI’13), IEEE, pp. 220-230. © IEEE, Los Alamitos, USA. 2013. 63. Miranda, L.C., Hornung, H.H., Pereira, R., and Baranauskas, M.C.C.. Exploring Adjustable Interactive Rings in Game Playing: Preliminary Results. In Proceedings of the 2nd International Conference on Design, User Experience and Usability / 15th International Conference on Human-Computer Interaction (DUXU/HCII’13), Springer, pp. 518–527. 2013. doi: 10.1007/978-3-642-39241-2_57. 64. Pereira, A.B.C., Souza Junior, G.N., Monteiro, D.C., Barros, E.S., Costa, H.P., Nascimento, P.A.A., Marques, L.B., Souza, D.G., Salgado, F.M., and Bessa, R.Q. A AIED game to help children with learning disabilities in literacy in the Portuguese language. In Anais do XI Simpósio Brasileiro de Jogos e Entretenimento Digital (SBGames’12), SBC, pp. 134–143. 2012. 65. Shindo, N. Dicionário de Nomes Japoneses. JBC, 2006. 66. Stubbs, K. Kana no senshi (kana warrior): a new interface for learning Japanese characters. In Proceedings of the Extended Abstracts on Human Factors in Computing Systems (CHI EA’03), ACM, pp. 894–895. 2003. doi: 10.1145/765891.766054. 67. Zichermann, G. and Cunningham, C.. Gamification by Design. O’Reilly. 2001. 68. Barcelos, T.S., Carvalho, T., Schimiguel, J. and Silveira, I.F. Comparative analysis of heuristics for digital games evaluation (in Portuguese). In Proc. IHC+CLIHC 2011, SBC, pp. 187-196. 2011. 69. Desurvire, H., Caplan, M. and Toth, J.A. Using heuristics to evaluate the playability of games. In Ext. Abstracts CHI 2004, ACM, pp. 1509-1512. 2004. 70. Federoff, M. Heuristics and usability guidelines for the creation and evaluation of fun in video games. Master thesis. Indiana University. 2002. 71. Følstad, A., Law, E. and Hornbæk, K. Analysis in practical usability evaluation: A survey study. In Proc. CHI 2012, ACM, pp. 2127-2136. 2012. 72. Grace, L., Castaneda, M. and Ducher, J. User testing of a language learning game for Mandarin Chinese. In Ext. Abstracts CHI 2012, ACM, pp. 865-868. 2012. 73. Grossman, T., Fitzmaurice, G. and Attar, R. A survey of software learnability: Metrics, methodologies and guidelines. In Proc. CHI 2009, ACM, pp. 649-658. 2009. 74. Hersh, M. and Leporini, B. Accessibility and usability of educational gaming environments for disabled students. In Proc. ICALT 2012, IEEE, pp. 752-753. 2012. 138

75. Hilbert, D.M. and Redmiles, D.F. Extracting usability information from user interface events. ACM Comput. Surv. 32, 4, pp. 384-421. 2000. 76. Ivory, M.Y. and Hearst, M.A. The state of the art in automating usability evaluation of user interfaces. ACM Comput. Surv. 33, 4, pp. 470-516. 2001. 77. Korhonen, H., Paavilainen, J. and Saarenpää, H. Expert review method in game evaluations: Comparison of two playability heuristic sets. In Proc. MindTrek 2009, ACM, pp. 74-81. 2009. 78. Lazareck, L.J., Farrell, D., Kostkova, P., Lecky, D.M., McNulty, C.A.M. and Weerasinghe, D. Learning by gaming - evaluation of an online game for children. In Proc. EMBC 2010, IEEE, pp. 2951-2952. 2010. 79. Liao, Y.H. and Shen, C.Y. Heuristic evaluation of digital game based learning: A case study. In Proc. DIGITEL 2012, IEEE, pp. 192-196. 2012. 80. Marciano, J.N., Ferreira, A.L.S., Correia, A.C.C., Miranda, L.C. and Miranda, E.E.C. Karuchā Ships Invaders: Cultural issues on the design/development of a Japanese CALL game made by/to Brazilians. In Proc. SBGames 2013, © SBC, pp. 172-180. 2013. 81. Mohamed, H. and Jaafar, A. Development and potential analysis of heuristic evaluation for educational computer game (PHEG). In Proc. ICCIT 2010, IEEE, pp. 222-227. 2010. 82. Mohamed-Omar, H., Yusoff, R. and Jaafar, A. Quantitive analysis in a heuristic evaluation for usability of educational computer game (UsaECG). In Proc. CAMP 2012, IEEE, pp.187-192. 2012. 83. Omar, H. and Jaafar, A. Heuristics evaluation in computer games. In Proc. CAMP 2010, IEEE, pp. 188-193. 2010. 84. Omar, H.M. and Jaafar, A. AHP_HeGES: Tools to evaluate usability of educational computer game (UsaECG). In Proc. i-USEr 2011, IEEE, pp. 73-76. 2011. 85. Omar, H.M. and Jaafar, A. Challenges in the evaluation of educational computer games. In Proc. ITSim 2010, IEEE, pp. 1-6. 2010. 86. Omar, H.M. and Jaafar, A. Conceptual framework for a heuristics based methodology for interface evaluation of educational games. Computer and Information Science 3, 2, pp. 211-219. 2010. 87. Omar, H.M. and Jaafar, A. Conceptual framework for a heuristics based methodology for interface evaluation of educational games. In Proc. ICCTD 2009, IEEE, pp. 594-598. 2009. 88. Omar, H.M. and Jaafar, A. Playability heuristics evaluation (PHE) approach for Malaysian educational games. In Proc. ITSim 2008, IEEE, pp. 1-7. 2008. 89. Omar, H.M., Ibrahim, R. and Jaafar, A. Methodology to evaluate interface of educational computer game. In Proc. ICPAIR 2011, IEEE, pp. 228-232. 2011. 90. Park, H.J. and Kim, S.B. Guidelines of serious game accessibility for the disabled. In Proc. ICISA 2013, IEEE pp. 1-3. 2013. 91. Pinelle, D., Wong, N. and Stach, T. Heuristic evaluation for games: Usability principles for video game design. In Proc. CHI 2008, ACM pp. 1453-1462. 2008. 139

92. Serrano, A., Marchiori, E.J., del Blanco, A., Torrente, J. and Fernandez-Manjon, B. A framework to improve evaluation in educational games. In Proc. EDUCON 2012, IEEE pp. 1-8. 2012. 93. Torrente, J., Marchiori, E.J., Vallejo-Pinto, J.A., Ortega-Moral, M., Moreno-Ger, P. and Fernandez-Manjon, B. Eyes-free interfaces for educational games. In Proc. SIIE 2012, IEEE pp. 1-6. 2012. 94. Yuan, B., Folmer, E. and Harris, F.C.Jr. Game accessibility: A survey. In Univers. Access Inf. Soc. vol. 10, issue 1, pp. 81-100. 2011. 95. Boyce, A.K., Campbell, A., Pickford, S., Culler, D. and Barnes, T. Experimental evaluation of BeadLoom game: how adding game elements to an educational tool improves motivation and learning. In Proceedings of the 16th Annual Joint Conference on Innovation and Technology in Computer Science Education, pp. 243-247. ACM, New York, NY, USA. 2011. 96. Franzwa, C., Tang, Y., and Johnson, A. Serious game design: motivating students through a balance of fun and learning. In Proceedings of the 5th International Conference on Games and Virtual Worlds for Serious Applications pp. 1-7. 2013. 97. Huang, D.W., Diefes-Dux, H., Imbrie, P.K., Daku, B., and Kallimani, J.G. Learning motivation evaluation for a computer-based instructional tutorial using ARCS model of motivational design. In 34th Annual Frontiers in Education pp. 30-36. 2004. 98. Keller, J. Strategies for stimulate the motivation to learn. In Performance & Instruction pp. 1-7. 1987. 99. Keller, J. Instructional Material Motivational Survey. In Florida State University, Unpublished documents. 1993. 100. Keller, J. How to integrate learner motivation planning into lesson planning: The ARCS model approach – Running head: Integrating motivation. In VII Semanario, pp. 1-17. 2000. 101. Münz, U., Schumm, P., Wiesebrock, A., and Allgower, F. Motivation and learning progress through educational games. In IEEE Transactions on Industrial Electronics, vol. 54, issue 6, pp. 3141- 3144. 2007. 102. Zaharias, P. A usability evaluation method for e-learning: focus on motivation to learn. In Extended Abstracts on Human Factors in Computing Systems, pp. 1571-1576. ACM, New York, NY, USA. 2006. 103. Malone, T.W. and Lepper, M.R. Making learning fun: a taxonomy of intrinsic motivations for learning. In Aptitude, Learning, and Instruction, vol. 3, pp. 223-253. 1987. 104. Tohidi, H. and Jabbari, M.M. The effects of motivation in education. In Procedia - Social and Behavioral Sciences, vol. 31, pp. 820-824. 2012. 105. Nielsen, J. Heuristic evaluation. In Nielsen, J., and Mack, R.L. (Eds.), Usability Inspection Methods, John Wiley & Sons, New York, NY. 1994. 140

106. Shellnut, B., Knowlton, A. and Savage, T.: Applying the ARCS model to the design and development of computer-based modules for manufacturing engineering courses. In: Educational Technology Research & Development, pp. 100–110. 1999. 107. Wongwiwatthananukit, S. and Popovich, N.G. Applying the ARCS model of motivational design to pharmaceutical education. In: American Journal of Pharmaceutical Education, pp. 188–196. 2000. 108. Marciano, J.N., Miranda, L.C. and Miranda, E.E.C.: Evaluation of motivational aspects in educational media and technology. In: 26th World Conference on Educational Media and Technology, pp. 780–787, © AACE 2014.