DOCSLIB.ORG

Redalyc.A Basic Language Technology Toolkit for Quechua

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Redalyc.A Basic Language Technology Toolkit for Quechua Procesamiento del Lenguaje Natural ISSN: 1135-5948 [email protected] Sociedad Española para el Procesamiento del Lenguaje Natural España Rios, Annette A Basic Language Technology Toolkit for Quechua Procesamiento del Lenguaje Natural, núm. 56, 2016, pp. 91-94 Sociedad Española para el Procesamiento del Lenguaje Natural Jaén, España Available in: http://www.redalyc.org/articulo.oa?id=515754423009 How to cite Complete issue Scientific Information System More information about this article Network of Scientific Journals from Latin America, the Caribbean, Spain and Portugal Journal's homepage in redalyc.org Non-profit academic project, developed under the open access initiative Procesamiento de Lenguaje Natural, Revista nº 56, marzo de 2016, pp 91-94 recibido 29-10-2015 revisado 17-12-2015 aceptado 28-01-2016 A Basic Language Technology Toolkit for Quechua Una colecci´onde herramientas b´asicas para el procesamiento autom´atico del Quechua Annette Rios Institute of Computational Linguistics University of Zurich [email protected] Resumen: Tesis escrita por Annette Rios en la Universidad de Z´urich bajo la di- recci´onde Prof. Dr. Martin Volk. La tesis fue defendida el 21 de septiembre de 2015 en la Universidad de Z´urich ante el tribunal formado por Prof. Dr. Martin Volk (Universidad de Z´urich, Departamento de Ling¨u´ısticaComputacional), Prof. Dr. Balthasar Bickel (Universidad de Z´urich, Departamento de Ling¨u´ısticaCompar- ativa) y Dr. Paul Heggarty (Instituto Max Planck para la Antropolog´ıaEvolutiva). La tesis obtuvo la calificaci´on`Summa cum Laude'. Palabras clave: Traducci´onautom´atica,an´alisismorfol´ogico,programa de con- cordancia, transductor de estados finitos, traducci´onautom´aticah´ıbrida,treebank, gram´aticade dependencias Abstract: Thesis written by Annette Rios under the supervision of Prof. Dr. Mar- tin Volk at the University of Zurich. The thesis defense was held at the University of Zurich on September 21, 2015 and was awarded `Summa Cum Laude'. The mem- bers of the committee were Prof. Dr. Martin Volk (University of Zurich, Institute of Computational Linguistics), Prof. Dr. Balthasar Bickel (University of Zurich, Department of Comparative Linguistics) and Dr. Paul Heggarty (Max Planck In- stitute for Evolutionary Anthropology). Keywords: Machine translation, morphological analysis, concordancer, finite state, hybrid MT, treebank, dependency grammar 1 Introduction also be used for closely related varieties. In this thesis, we describe the development The main focus of this work lies on the of several natural language processing tools implementation of a machine translation sys- and resources for the Andean language Cuzco tem for the language pair Spanish-Cuzco Quechua as part of the SQUOIA project at Quechua. Since the target language Quechua the University of Zurich. is not only a non-mainstream language in Quechua is a group of closely related the field of computational linguistics, but languages, spoken by 8-10 million people in also typologically quite different from the Peru, Bolivia, Ecuador, Southern Colombia source language Spanish, several rather un- and the North-Western parts of Argentina. usual problems became evident, and we had Although Quechua is often referred to to find solutions in order to deal with them. as a `language' and its local varieties as Therefore, the first part of this thesis presents `dialects', Quechua represents a language monolingual tools and resources that are not family, comparable in depth to the Romance directly related to machine translation, but or Slavic languages (Adelaar and Muysken, are nevertheless indispensable. 2004).Mutual intelligibility, especially be- All resources and tools are freely available tween speakers of distant dialects, is not from the project's website.1 always given. The applications described Apart from the scientific interest in devel- in this thesis were developed for Cuzco oping tools and applications for a language Quechua, but some applications, such as the morphology tools and the spell checker, can 1https://github.com/ariosquoia/squoia ISSN 1135-5948 © 2016 Sociedad Española para el Procesamiento del Lenguaje Natural Annette Rios that is typologically distant from the main- directly related to machine translation, au- stream languages in computational linguis- tomatic processing of Quechua morphology tics, we hope that the various resources pre- provides the necessary resources for impor- sented in this thesis will be useful not only tant parts of the translation system. for language learners and linguists, but also Moreover, the project involved the cre- to Quechua speakers who want to use modern ation of a parallel treebank in all 3 languages. technology in their native language. While the Spanish and German parts of the treebank were finished within the first year 2 Structure of Thesis of the project, building the Quechua tree- The thesis is structured as follows: bank took considerably longer: before the ac- In the first chapter, we set the broader tual annotation process started, the texts had context for the development of NLP tools for to be translated into Quechua. Additionally, a low-resource language and we give a short we had to design an annotation scheme from overview on the characteristics and the dis- scratch and set up pre-processing and anno- tribution of the Quechua languages. tation tools. Chapter 2 explains the morphological A by-product of the resulting parallel cor- structures in Quechua word formation and pus is the Spanish-Quechua version of Bil- how we deal with them in technological ap- ingwis, a web tool that allows to search for plications, this includes morphological anal- translations in context in word-aligned texts. ysis, disambiguation, automatic text normal- Generally speaking, the monolingual tools ization and spell checking. and resources described in the first part of Chapter 3 summarizes the treebanking this thesis are necessary to build the multi- process and how existing tools were adapted lingual applications of the second part. for the syntactic annotation of Quechua texts 3 Contributions with dependency trees. Chapter 4 describes how Bilingwis, an on- The main contributions of this thesis are as line service for searching translations in par- follows: allel, word-aligned texts, was adapted to the language pair Spanish-Quechua. • We built a hybrid machine translation Chapter 5 describes the implementation of system that can translate Spanish text the hybrid machine translation system for the into Cuzco Quechua. The core system language pair Spanish-Quechua, with a spe- is a classical rule-based transfer engine, cial focus on resolving morphological ambi- however, several statistical modules are guities. included for tasks that cannot be re- The main part of this work was the de- solved reliably with rules. velopment of the translation system Spanish- • We have created an extensive finite Cuzco Quechua. However, the special situa- state morphological analyzer for South- tion of the target language Quechua as a non- ern Quechua that achieves high cover- mainstream language in computational lin- age. The analyzer consists of a set cas- guistics and as a low-prestige language in so- caded finite state transducers, where the ciety resulted in many language specific prob- last transducer is a guesser that can an- lems that had to be solved along the way. alyze words with unknown roots. Fur- For instance, the wide range of different thermore, we included the Spanish lem- orthographies used in written Quechua is a mas from the FreeLing library2 into the problem for any statistical text processing, analyzer in order to recognize the nu- such as the training of a language model to merous Spanish loan words in Quechua rank different translation options. Thus, in texts (words that consist of a Spanish order to get a statistical language model, we root with Quechua suffixes). had to first find a way to normalize Quechua • We implemented a text normaliza- texts automatically. tion pipeline that automatically rewrites Furthermore, the resulting normalization Quechua texts in different orthographies pipeline can be adapted for spell checking or dialects to the official Peruvian stan- with little effort. For this reason, an entire dard orthography. Additionally, we cre- chapter of this thesis is dedicated to the treat- ment of Quechua morphology: although not 2http://nlp.lsi.upc.edu/freeling/ 92 A Basic Language Technology Toolkit for Quechua ated a slightly adapted version that translation system was trained on normalized can be used as spell checker back-end, morphemes instead of word forms, in order to in combination with a plug-in for the mitigate data sparseness. open-source productivity suite LibreOf- Apart from the rich morphology, Quechua fice/OpenOffice.3 has several characteristics that have rarely (or not at all) been dealt with in machine • We built a Quechua dependency tree- translation. One of the most important issues bank of about 2000 annotated sentences, concerns verb forms in subordinated clauses: that provided not only training data for while Spanish has mostly finite verbs in sub- some of the translation modules, but ordinated clauses that are marked for tense, also served as a source of verification, aspect, modality and person, Quechua often since it allows to observe the distribu- has nominal forms that vary according to tion of certain syntactic and morpholog- the clause type. Furthermore, Quechua uses ical structures. Furthermore, we trained switch-reference as a device of clause linkage, a statistical parser on the treebank and while in Spanish, co-reference of subjects is thus have now a complete pipeline to unmarked and pronominal subjects are usu- morphologically analyze, disambiguate ally omitted (`pro-drop'). This leads to am- and then parse Quechua texts. biguities when Spanish text is translated into Quechua. 4 Conclusions and Outlook Another special case are relative clauses: We have created tools and applications for a the form of the nominalized verb in the language with very limited resources: While Quechua relative clause depends on whether printed Quechua dictionaries and grammars the head noun is the semantic agent of the rel- exist, some of them quite outdated, digital ative clause.
Load more

Recommended publications
  • Preparation and Exploitation of Bilingual Texts Dusko Vitas, Cvetana Krstev, Eric Laporte
    Preparation and exploitation of bilingual texts Dusko Vitas, Cvetana Krstev, Eric Laporte To cite this version: Dusko Vitas, Cvetana Krstev, Eric Laporte. Preparation and exploitation of bilingual texts. Lux Coreana, 2006, 1, pp.110-132. hal-00190958v2 HAL Id: hal-00190958 https://hal.archives-ouvertes.fr/hal-00190958v2 Submitted on 27 Nov 2007 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Preparation and exploitation of bilingual texts Duško Vitas Faculty of Mathematics Studentski trg 16, CS-11000 Belgrade, Serbia Cvetana Krstev Faculty of Philology Studentski trg 3, CS-11000 Belgrade, Serbia Éric Laporte Institut Gaspard-Monge, Université de Marne-la-Vallée 5, bd Descartes, 77454 Marne-la-Vallée CEDEX 2, France Introduction A bitext is a merged document composed of two versions of a given text, usually in two different languages. An aligned bitext is produced by an alignment tool or aligner, that automatically aligns or matches the versions of the same text, generally sentence by sentence. A multilingual aligned corpus or collection of aligned bitexts, when consulted with a search tool, can be extremely useful for translation, language teaching and the investigation of literary text (Veronis, 2000).
    [Show full text]
  • The Iafor European Conference Series 2014 Ece2014 Ecll2014 Ectc2014 Official Conference Proceedings ISSN: 2188-1138
    the iafor european conference series 2014 ece2014 ecll2014 ectc2014 Official Conference Proceedings ISSN: 2188-1138 “To Open Minds, To Educate Intelligence, To Inform Decisions” The International Academic Forum provides new perspectives to the thought-leaders and decision-makers of today and tomorrow by offering constructive environments for dialogue and interchange at the intersections of nation, culture, and discipline. Headquartered in Nagoya, Japan, and registered as a Non-Profit Organization 一般社( 団法人) , IAFOR is an independent think tank committed to the deeper understanding of contemporary geo-political transformation, particularly in the Asia Pacific Region. INTERNATIONAL INTERCULTURAL INTERDISCIPLINARY iafor The Executive Council of the International Advisory Board IAB Chair: Professor Stuart D.B. Picken IAB Vice-Chair: Professor Jerry Platt Mr Mitsumasa Aoyama Professor June Henton Professor Frank S. Ravitch Director, The Yufuku Gallery, Tokyo, Japan Dean, College of Human Sciences, Auburn University, Professor of Law & Walter H. Stowers Chair in Law USA and Religion, Michigan State University College of Law Professor David N Aspin Professor Emeritus and Former Dean of the Faculty of Professor Michael Hudson Professor Richard Roth Education, Monash University, Australia President of The Institute for the Study of Long-Term Senior Associate Dean, Medill School of Journalism, Visiting Fellow, St Edmund’s College, Cambridge Economic Trends (ISLET) Northwestern University, Qatar University, UK Distinguished Research Professor of Economics,
    [Show full text]
  • Kernerman Kdictionaries.Com/Kdn DICTIONARY News the European Network of E-Lexicography (Enel) Tanneke Schoonheim
    Number 22 ● July 2014 Kernerman kdictionaries.com/kdn DICTIONARY News The European Network of e-Lexicography (ENeL) Tanneke Schoonheim On October 11th 2013, the kick-off meeting of the European production and reception of dictionaries. The internet offers Network of e-Lexicography (ENeL) project took place in entirely new possibilities for developing and presenting Brussels. This meeting was the outcome of an idea ventilated dictionary information, such as with the integration of sound, a year and a half earlier, in March 2012 in Berlin, at the maps or video, and various novel ways of interacting with European Workshop on Future Standards in Lexicography. dictionary users. For editors of scholarly dictionaries the new The workshop participants then confirmed the imperative to medium is not only a source of inspiration, it also generates coordinate and harmonise research in the field of (electronic) new and serious challenges that demand cooperation and lexicography across Europe, namely to share expertise relating standardization on various levels: to standards, discuss new methodologies in lexicography that a. Through the internet scholarly dictionaries can potentially fully exploit the possibilities of the digital medium, reflect on reach large audiences. However, at present scholarly the pan-European nature of the languages of Europe and attain dictionaries providing reliable information are often not easy a wider audience. to find and are hard to decode for a non-academic audience; A proposal was written by a team of researchers from
    [Show full text]
  • TANGO: Bilingual Collocational Concordancer
    TANGO: Bilingual Collocational Concordancer Jia-Yan Jian Yu-Chia Chang Jason S. Chang Department of Computer Inst. of Information Department of Computer Science System and Applictaion Science National Tsing Hua National Tsing Hua National Tsing Hua University University University 101, Kuangfu Road, 101, Kuangfu Road, 101, Kuangfu Road, Hsinchu, Taiwan Hsinchu, Taiwan Hsinchu, Taiwan [email protected] [email protected] [email protected] du.tw on elaborated statistical calculation. Moreover, log Abstract likelihood ratios are regarded as a more effective In this paper, we describe TANGO as a method to identify collocations especially when the collocational concordancer for looking up occurrence count is very low (Dunning, 1993). collocations. The system was designed to Smadja’s XTRACT is the pioneering work on answer user’s query of bilingual collocational extracting collocation types. XTRACT employed usage for nouns, verbs and adjectives. We first three different statistical measures related to how obtained collocations from the large associated a pair to be collocation type. It is monolingual British National Corpus (BNC). complicated to set different thresholds for each Subsequently, we identified collocation statistical measure. We decided to research and instances and translation counterparts in the develop a new and simple method to extract bilingual corpus such as Sinorama Parallel monolingual collocations. Corpus (SPC) by exploiting the word- We also provide a web-based user interface alignment technique. The main goal of the capable of searching those collocations and its concordancer is to provide the user with a usage. The concordancer supports language reference tools for correct collocation use so learners to acquire the usage of collocation.
    [Show full text]
  • Aconcorde: Towards a Proper Concordance for Arabic
    aConCorde: towards a proper concordance for Arabic Andrew Roberts, Dr Latifa Al-Sulaiti and Eric Atwell School of Computing University of Leeds LS2 9JT United Kingdom {andyr,latifa,eric}@comp.leeds.ac.uk July 12, 2005 Abstract Arabic corpus linguistics is currently enjoying a surge in activity. As the growth in the number of available Arabic corpora continues, there is an increased need for robust tools that can process this data, whether it be for research or teaching. One such tool that is useful for both groups is the concordancer — a simple tool for displaying a specified target word in its context. However, obtaining one that can reliably cope with the Arabic lan- guage had proved extremely difficult. Therefore, aConCorde was created to provide such a tool to the community. 1 Introduction A concordancer is a simple tool for summarising the contents of corpora based on words of interest to the user. Otherwise, manually navigating through (of- ten very large) corpora would be a long and tedious task. Concordancers are therefore extremely useful as a time-saving tool, but also much more. By isolat- ing keywords in their contexts, linguists can use concordance output to under- stand the behaviour of interesting words. Lexicographers can use the evidence to find if a word has multiple senses, and also towards defining their meaning. There is also much research and discussion about how concordance tools can be beneficial for data-driven language learning (Johns, 1990). A number of studies have demonstrated that providing access to corpora and concordancers benefited students learning a second language.
    [Show full text]
  • Panacea D6.1
    SEVENTH FRAMEWORK PROGRAMME THEME 3 Information and communication Technologies PANACEA Project Grant Agreement no.: 248064 Platform for Automatic, Normalized Annotation and Cost-Effective Acquisition of Language Resources for Human Language Technologies D6.1 Technologies and Tools for Lexical Acquisition Dissemination Level: Public Delivery Date: July 16th 2010 Status – Version: Final Author(s) and Affiliation: Laura Rimell (UCAM), Anna Korhonen (UCAM), Valeria Quochi (ILC-CNR), Núria Bel (UPF), Tommaso Caselli (ILC-CNR), Prokopis Prokopidis (ILSP), Maria Gavrilidou (ILSP), Thierry Poibeau (UCAM), Muntsa Padró (UPF), Eva Revilla (UPF), Monica Monachini(CNR-ILC), Maurizio Tesconi (CNR-IIT), Matteo Abrate (CNR-IIT) and Clara Bacciu (CNR-IIT) D6.1 Technologies and Tools for Lexical Acquisition This document is part of technical documentation generated in the PANACEA Project, Platform for Automatic, Normalized Annotation and Cost-Effective Acquisition (Grant Agreement no. 248064). This documented is licensed under a Creative Commons Attribution 3.0 Spain License. To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/es/. Please send feedback and questions on this document to: [email protected] TRL Group (Tecnologies dels Recursos Lingüístics), Institut Universitari de Lingüística Aplicada, Universitat Pompeu Fabra (IULA-UPF) D6.1 – Technologies and Tools for Lexical Acquisition Table of contents Table of contents ..........................................................................................................................
    [Show full text]
  • Translate's Localization Guide
    Translate’s Localization Guide Release 0.9.0 Translate Jun 26, 2020 Contents 1 Localisation Guide 1 2 Glossary 191 3 Language Information 195 i ii CHAPTER 1 Localisation Guide The general aim of this document is not to replace other well written works but to draw them together. So for instance the section on projects contains information that should help you get started and point you to the documents that are often hard to find. The section of translation should provide a general enough overview of common mistakes and pitfalls. We have found the localisation community very fragmented and hope that through this document we can bring people together and unify information that is out there but in many many different places. The one section that we feel is unique is the guide to developers – they make assumptions about localisation without fully understanding the implications, we complain but honestly there is not one place that can help give a developer and overview of what is needed from them, we hope that the developer section goes a long way to solving that issue. 1.1 Purpose The purpose of this document is to provide one reference for localisers. You will find lots of information on localising and packaging on the web but not a single resource that can guide you. Most of the information is also domain specific ie it addresses KDE, Mozilla, etc. We hope that this is more general. This document also goes beyond the technical aspects of localisation which seems to be the domain of other lo- calisation documents.
    [Show full text]
  • Corpus Linguistics and Data-Driven Learning: a Critical Overview
    Published in Bulletin VALS-ASLA 97, 97-118, 2013 which should be used for any reference to this work Corpus linguistics and data-driven learning: a critical overview Alex BOULTON Université de Lorraine et Crapel, ATILF (UMR 7118) 44 avenue de la libération, 54000 Nancy, France [email protected] Henry TYNE Université de Perpignan Via Domitia et VECT (EA 2983) 52 avenue Paul Alduy, 66000 Perpignan, France [email protected] L'utilisation de corpus s'est avérée intéressante dans de nombreux domaines depuis plus de vingt ans, notamment en didactique des langues. Un état de l'art des études publiées dans ce domaine nous permettra de constater que si les résultats sont en général assez positifs (l'apprentissage a lieu), l'exploitation de corpus reste une activité marginale. Si, en général, de telles recherches en didactique des langues ne donnent pas toujours des résultats fulgurants, on pourrait faire l’hypothèse que les approches sont plus ou moins adaptées à certains types d'apprenants, ou à certains types d'activité, et qu'elles contribuent au bagage général des outils et des techniques disponibles. Qui plus est, avec une approche sur corpus, on pourrait de même croire que les étudiants deviennent tout simplement des apprenants plus conscients de leur apprentissage grâce au travail effectué. Nous suggérons que l'apport des corpus pour l'apprentissage des langues offre ainsi de nombreuses possibilités. Mots-clés: Corpus, apprentissage sur corpus, pratiques ordinaires d'enseignement-apprentissage, processus, TIC 1. Introduction Corpus linguistics is not simply a recondite field of research within linguistics: it affords practical methodologies and tools to further the study of all aspects of language use (e.g.
    [Show full text]
  • From a Bilingual Concordancer to a Translation Finder Julien Bourdaillet, Stéphane Huet, Philippe Langlais, Guy Lapalme
    TransSearch: from a bilingual concordancer to a translation finder Julien Bourdaillet, Stéphane Huet, Philippe Langlais, Guy Lapalme To cite this version: Julien Bourdaillet, Stéphane Huet, Philippe Langlais, Guy Lapalme. TransSearch: from a bilingual concordancer to a translation finder. Machine Translation, Springer Verlag, 2010, 24 (3-4), pp.241-271. 10.1007/s10590-011-9089-6. hal-02021930 HAL Id: hal-02021930 https://hal.archives-ouvertes.fr/hal-02021930 Submitted on 26 Feb 2019 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. 2 J. Bourdaillet, S. Huet, P. Langlais, G. Lapalme Figure 1. English sentences, along with their French translation, where the query in keeping with occurs in the current version of the bilingual concordancer Trans- Search. The query is highlighted in the English sentences, but the user has to search its translations in the French sentences. This makes the discovery of the different translation possibilities more difficult (adapt´ees`a, conforme aux, d´etonne and conform´ement`a for the four displayed sentences), and forces the user to scroll down the page in order to read other sentences while searching for further translation material. This paper describes the enhancement of the web-based bilingual concordancer TransSearch1 (Macklovitch et al., 2000).
    [Show full text]
  • Computer Aided Translation Advances and Challenges
    Computer Aided Translation Advances and Challenges Philipp Koehn 7 August 2016 Philipp Koehn Computer Aided Translation 7 August 2016 Overview 1 • A practical introduction: the casmacat workbench • Postediting • Types of assistance • Logging, eye tracking and user studies • Implementation details of the casmacat workbench : Philipp Koehn Computer Aided Translation 7 August 2016 2 part I casmacat workbench Philipp Koehn Computer Aided Translation 7 August 2016 3 casmacat workbench Philipp Koehn Computer Aided Translation 7 August 2016 CASMACAT Project 2011-2014 4 • Cognitive studies of translators leading to insights into interface design ! better understanding of translator needs • Workbench with novel types of assistance to human translators { interactive translation prediction { interactive editing and reviewing { adaptive translation models ! better tools for translators • Demonstration of effectiveness in field tests with professional translators ! increased translator productivity Philipp Koehn Computer Aided Translation 7 August 2016 Core Modes 5 Philipp Koehn Computer Aided Translation 7 August 2016 Postediting Modes 6 Philipp Koehn Computer Aided Translation 7 August 2016 GUI Features 7 Philipp Koehn Computer Aided Translation 7 August 2016 Postediting Interface 8 • Source on left, translation on right • Context above and below Philipp Koehn Computer Aided Translation 7 August 2016 Confidence Measures 9 • Sentence-level confidence measures ! estimate usefulness of machine translation output • Word-level confidence measures ! point
    [Show full text]
  • The Universal Dependencies Treebank of Spoken Slovenian
    The Universal Dependencies Treebank of Spoken Slovenian Kaja Dobrovoljc1, Joakim Nivre2 1Institute for Applied Slovene Studies Trojina, Ljubljana, Slovenia 1Department of Slovenian Studies, Faculty of Arts, University of Ljubljana 2Department of Linguistics and Philology, Uppsala University [email protected], joakim.nivre@lingfil.uu.se Abstract This paper presents the construction of an open-source dependency treebank of spoken Slovenian, the first syntactically annotated collection of spontaneous speech in Slovenian. The treebank has been manually annotated using the Universal Dependencies annotation scheme, a one-layer syntactic annotation scheme with a high degree of cross-modality, cross-framework and cross-language interoper- ability. In this original application of the scheme to spoken language transcripts, we address a wide spectrum of syntactic particularities in speech, either by extending the scope of application of existing universal labels or by proposing new speech-specific extensions. The initial analysis of the resulting treebank and its comparison with the written Slovenian UD treebank confirms significant syntactic differences between the two language modalities, with spoken data consisting of shorter and more elliptic sentences, less and simpler nominal phrases, and more relations marking disfluencies, interaction, deixis and modality. Keywords: dependency treebank, spontaneous speech, Universal Dependencies 1. Introduction actually out-performs state-of-the-art pipeline approaches (Rasooli and Tetreault, 2013; Honnibal and Johnson, 2014). It is nowadays a well-established fact that data-driven pars- Such heterogeneity of spoken language annotation schemes ing systems used in different speech-processing applica- inevitably leads to a restricted usage of existing spoken tions benefit from learning on annotated spoken data, rather language treebanks in linguistic research and parsing sys- than using models built on written language observation.
    [Show full text]
  • Arxiv:2105.02877V1 [Cs.CV] 6 May 2021 Tles Corresponding to the Audio Content
    Aligning Subtitles in Sign Language Videos Hannah Bull1* Triantafyllos Afouras2∗ Gul¨ Varol2;3 Samuel Albanie2 Liliane Momeni2 Andrew Zisserman2 1 LISN, Univ Paris-Saclay, CNRS, France 2 Visual Geometry Group, University of Oxford, UK 3 LIGM, Ecole´ des Ponts, Univ Gustave Eiffel, CNRS, France [email protected]; fafourast,gul,albanie,liliane,[email protected] https://www.robots.ox.ac.uk/˜vgg/research/bslalign/ Overlooked by a small hill known as Leopard the mother must make sure To keep her cubs alive in this dangerous neighbourhood, Saudio Rock. they stay hidden. Sgt Overlooked by a small hill known as Leopard Rock. the mother must make sure they stay hidden. To keep her cubs alive in this dangerous neighbourhood, Time 14:12 14:14 14:16 14:18 14:20 14:22 14:24 14:26 14:28 14:30 Figure 1: Subtitle alignment: We study the task of aligning subtitles to continuous signing in sign language interpreted TV broadcast data. The subtitles in such settings usually correspond to and are aligned with the audio content (top: audio subtitles, Saudio) but are unaligned with the accompanying signing (bottom: Ground Truth annotation of the signing corresponding to the subtitle, Sgt). This is a very challenging task as (i) the order of subtitles varies between spoken and sign languages, (ii) the duration of a subtitle differs considerably between signing and speech, and (iii) the signing corresponds to a translation of the speech as opposed to a transcription. Abstract evaluations, we show substantial improvements over exist- ing alignment baselines that do not make use of subtitle text The goal of this work is to temporally align asyn- embeddings for learning.
    [Show full text]