• Web-based eBooks with activities: internationalization of Natural Language Processing
The Client : Javna agencija za raziskovalno dejavnost RS
Project type: Bilateral projects
Project duration: 2018 - 2019
  • Description

Research area From an educational standpoint, Slovenia, like many modern countries, has a recreational reading problem with its adolescents. The many adolescents who choose not to read do NOT learn valuable reading skills that are vital for academic progress and later professional success. Many high school students choose not to read recreationally. This is a serious problem in terms of the development of their reading skills, academic success and ultimately their professional lives. Although recreational reading is vital to reading skill (Share, 2008), more than half of adolescents in Western industrialized countries do not read recreationally (OECD, 2010), resulting in a downward causal spiral of disinterest in reading, decreasing time spent reading, and a decline in reading skills (Mol & Bus, 2011). This diminishes academic and professional success (Schlackman, Gottfried, & Boutin-Martinez, 2015; Notten, 2011; OECD, 2010; Taylor, 2013). Web-based eBooks with activities or games (IMapBook project) at University of South Florida were designed to increase recreational reading by bringing books into the digital recreational space of adolescents. These web-based books with social interaction and games involving conversations with characters within books, have been used with great success in a number of educational projects and in the research collaborations in the United States, Netherlands and China. With IMapBooks, kids read a chapter of text, and then come to a game that can only be win by comprehending key points from the chapter. They need to win the game to read on. For example, conversations with characters leverage Natural Language Processing (NLP), which makes reader conversations with fictional characters feel natural. In other words, NLP lets readers type or speak words to their computer screen, and the character answers them in a natural way. Goals IMapBooks mission is to increase engagement in reading and writing by upgrading the digital book technology to include online games, conversations with characters in the stories and social interaction. The goals of the cooperation between the Slovenian and the US research groups will have the following effects: SCIENTIFIC: To extend the existing IMapBooks goals, we are working on adjustments of NLP techniques for the Slovene language with embedded games involving digital chat conversations with fictional characters & chats between readers, to (a) promote adolescent recreational reading in Slovenia, and (b) use as test cases for NLP in Slovene. Although 315 million people speaking Slavic languages, they are NOT well supported by NLP tools. On the other side English, as a de-facto world language, is already quite well supported by NLP tools. As a final result of this project, the Slovene language will be richer in having linguistic corpora and in having developed new NLP techniques for better equipment of the Slovene language with new natural language technologies. SOCIAL: We strive to deepen the user experience of reading a book, for those who dont enjoy reading, by leverage those things that avid readers enjoy about reading. For instance, as people read novels, they alternate times reading with time not reading. While not reading, readers simmer in the atmosphere of the book, in the characters, in the conflict situation, and they try to predict what will happen (larger predictive inferences), and engage in reflective thinking about the book. Games within a book can deepen the reading experience in a similar way (both while reading and while not reading), by deepening the readers knowledge and involvement with the world, adding depth, providing more clues for inferencing (without giving the outcome away). Organization & efficiency The IMapBook research group is a part of a larger Department of Educational & Psychological Studies at the University of South Florida. It is active since 2007 and has invented a new book technology for which they secured grants for financial support (National Science Foundation, USA), conducted research in a number of countries, secured a US patent, improved high school content learning such as climate change science and published a series of peer-reviewed journal articles. Current core personnel include professors, post-doc students, doctoral students and graduate assistants. The Laboratory for Data Technologies research group is comprised of many researchers that work in text analysis, word processing, network and general data analysis. Thus, we estimate that collaboration with the US research group that we have recently started collaborating with, and which perfectly supplements our knowledge, we can successfully carry out the proposed project. In addition, the group also consists of junior researchers at the faculty and those from industry (postgraduate students at the University of Ljubljana), which present the basis for the future research in the field of the proposal. Contribution The components of proposed IMapBooks for Slovene and a NLP library will improve the English authoring system and completely support Slovene language, which represents the main contribution of the project. In addition, the proposed project makes it easier to explore this highly interdisciplinary field, where collaboration is indeed needed. Closer cooperation with the extremely wellestablished US research group presents an important link that can lead to numerous joint projects in the future. Significant results and new findings, which will result from project participation, will be presented at well recognized international scientific conferences and published in scientific journals. The joint work will also greatly help the existing IMapBook project, also by expanding into another family of languages. It would provide some new recreational reading tools for adolescents, improve existing authoring tools and add new types of embedded games. Thus, the benefits are mutual and synergistic.