Automated Online Communication: How to Interact in the 21st Century Dissertation by Zaenal Akbar submitted to the Faculty of Mathematics, Computer Science and Physics of the University of Innsbruck in partial fulfillment of the requirements for the degree of Doctor of Philosophy (PhD) Supervisor: Ass.-Prof. Dr. Anna Fensel Innsbruck, June 2018 © Copyright Zaenal Akbar The material in this publication is protected by copyright law. Year: 2018 Title: Automated Online Communication: How to Interact in the 21st Century Author: Zaenal Akbar Acknowledgements I would like to thank my supervisor Ass.-Prof. Dr. Anna Fensel for giving me the opportunity to work under her supervision. Anna provided me with valuable advice and guidance such that I can complete my research works and finish this dissertation. Further, I am also deeply grateful to the Chair of STI Innsbruck, Univ.-Prof. Dr. Dieter Fensel, who has provided me with a stimulating working environment at STI Innsbruck. Dieter has supported my works from the beginning, and this dissertation would never be finished without his valuable inputs and leadership. I am always looking forward to collaborating with both of them in the future. Many thanks go to my colleagues from STI Innsbruck, which whom I had great work- ing experiences and had co-authored papers. They had outstanding contributions to the im- provement of my works with their valuable ideas and feedback. I have carried out the re- search reported in this dissertation at the Online Communication Working Group1, Seman- tic Technology Institute (STI) Innsbruck2, Department of Computer Science3, University of Innsbruck4, Innsbruck, Austria. During the research, I have received support from various re- search projects funded by the European Union and the government of Austria. These projects included PlanetData5, TourPack6, and EuTravel7. Last but not least, I would like to thank my family for their never-ending support and encouragements. Especially to my wife Andi Ririn Noviarti, I am very grateful for her patient and loving support, for her commitment to raising a small family in Innsbruck during my research and the work on my dissertation. 1http://oc.sti2.at 2http://sti2.at 3http://informatik.uibk.ac.at/ 4http://uibk.ac.at 5https://www.planet-data.eu/ 6http://tourpack.sti2.at/ 7http://www.eutravelproject.eu/ iv Acknowledgements Abstract The Internet has introduced multiple online communication channels, widening the capa- bility of organizations to communicate and engage with the broadest audience possible, ex- panding their marketing outreach. For organizations in the tourism industry, the Internet is the primary source for travel planning where travel inspirations are mainly originating from social networking, videos/photos sharing sites, as well as search engines. Being present on- line is a must, but being present and active on multiple online communication channels is even more crucial. We identified three main challenges of a multi-channel online communication system, namely: (i) scalable communication, where the number of information sources and commu- nication channels is exponentially growing, (ii) bilateral communication, where communi- cation has been developed into multiple directions, (iii) personalized communication, where communication becomes more specific. To overcome these challenges, we proposed a so- lution, so-called Multi-Channel Online Communication framework. The framework utilizes semantic web technologies for: (i) information management, which represents scattered in- formation into a common information model, (ii) channel management, which represents a variety of communication activities as a common communication model, (iii) matchmaking mechanism, which aligns both information and communication models through rules, a form of knowledge representation. We applied the framework to organizations in the tourism industry, where we evaluated its impacts on the organizations in two aspects, namely: the increment of online visibility and the reduction of manually performed tasks. For online visibility, we assessed two sources of online visibility, namely search engines and social media. We expected richer search results on search engines that could drive more visitors. More traffic from multiple social media channels was expected as well. For the reduction of manually performed tasks, we assessed three activities that are still mainly performed by humans, namely content integration, content distribution, and user engagements integration. The results of the evaluation are encouraging, a combination of semantic annotation and multi-channel communication solutions is capable of doubling the online visibility of organizations while reducing the manually performed tasks by half. Supervisor: Ass.-Prof. Dr. Anna Fensel – University of Innsbruck, Austria Co-supervisor: assoz. Prof. Dr. Radu Prodan – University of Innsbruck, Austria vi Abstract List of Relevant Publications Peer-reviewed Papers Paper 1. Akbar, Z., García, J. M., Toma, I., and Fensel, D. (2014). On using semantically- aware rules for efficient online communication. In Proceedings of the 8th Inter- national Web Rule Symposium (RuleML 2014), Prague, Czech Republic, August 18-20, 2014, pages 37-51. Springer International Publishing. https://doi.org/10.1007/978-3-319-09870-8_3 Abstract: The ever growing number of communication channels not only enables a broader outreach for organizations, but also makes it more difficult for them to manage a very large number of channels and adapted content efficiently. Thus, finding the right channels to disseminate some content and adapting this content to specific channel requirements are real challenges for sharing information both efficiently and effectively. In this work, we present a rule-based system that ad- dresses these challenges by decoupling the information to be shared from the ac- tual channels where it is published. We propose semantic models to characterize and integrate various information sources and channels. A set of independent rules then interrelates these models, specifying the concrete publication workflow and content adaptation required. Furthermore, we evaluate our rule-based system using two different use cases, discussing the added value that the defined rules provide to this scenario and how they contribute to overcoming the identified challenges effectively. Paper 2. Akbar, Z., García, J. M., Toma, I., and Fensel, D. (2014). A semantic-based plat- form for efficient online communication. In Proceedings of the 13th International Semantic Web Conference - Posters & Demonstrations Track (ISWC-PD ’14), Riva del Garda, Italy, October 19-23, 2014, pages 337-340. CEUR-WS.org. http://ceur-ws.org/Vol-1272/paper_84.pdf Abstract: To achieve an effective and efficient way of disseminating information to ever growing communication channels, we propose an approach that separates the information and communication channels and interlinks them with an interme- diary component. The separation enables various dimensions to reuse the infor- mation and communication channels in transactional communication. In this paper we introduce our online communication platform, which is comprised of several components. The important roles of semantic web technologies to the platform are explained in detail, including a use case to show the contributions of semantic web in supporting the effectiveness and efficiency of information dissemination. viii List of Relevant Publications Paper 3. Akbar, Z., García, J. M., Toma, I., and Fensel, D. (2015). Measuring the impact of content adaptation on social media engagement. In Proceedings of the 2nd Euro- pean Conference on Social Media (ECSM 2015), Porto, Portugal, July 9-10, 2015, pages 1-10. ACPI. Abstract: The ultimate goal of social media marketing is to reach and engage with the widest audience possible. Usually, the content published on social media chan- nels must be adapted in order to meet the requirements of the target channels. This process, known as content adaptation, requires human knowledge to determine how the content should be prepared to meet the channel requirements, and to spread it across the channel as fast as possible. Extracting this knowledge for content adap- tation is a challenging task due to the constant evolution of the variety of human experiences and social media channels. In this paper, we show how we acquire and learn this knowledge as our first step towards content adaptation automation. We determine the social media posts properties that could be affected by the content adaptation. Those properties are: (i) presentation property where a post can be seen as a collection of terms, (ii) hypermedia property where a post can be seen as a col- lection of media connected by hyperlinks, and (iii) named-entity property where a post can be seen as a collection of real world entities. Based on these properties, the most engaged posts are selected and their similarity degrees to their original sources are computed and compared. From the similarity degree comparison, we can devise recommendations on how to obtain the highest engagement on a particular social channel by fine-tuning the content adaptation on each property. Paper 4. Akbar, Z., Toma, I., and Fensel, D. (2016). Optimizing the publication flow of touristic service providers on multiple social media channels. In Proceedings of the 2016 International Conference on Information and Communication Technolo- gies in Tourism (ENTER 2016), Bilbao, Spain, February 2-5, 2016, pages 211-224. Springer