Enhancing Access to Online Education: Quality Machine Translation of MOOC Content

Enhancing Access to Online Education: Quality Machine Translation of MOOC Content Valia Kordoni1, Antal van den Bosch2, Katia Kermanidis3, Vilelmini Sosoni3, Kostadin Cholakov1, Iris Hendrickx2, Matthias Huck4, Andy Way5 1Department of English and American Studies, Humboldt University, Berlin, Germany 2Centre for Language Studies, Radboud University, Nijmegen, Netherlands 3Ionian University, Corfu, Greece 4School of Informatics, University of Edinburgh, UK 5ADAPT Centre, School of Computing, Dublin City University, Ireland evangelia.kordoni@anglistik.hu-berlin.de, a.vandenbosch@let.ru.nl, kerman@ionio.gr, vilelmini@hotmail.com, cholakov@anglistik.hu-berlin.de, i.hendrickx@let.ru.nl, mhuck@inf.ed.ac.uk, away@computing.dcu.ie Abstract The present work is an overview of the TraMOOC (Translation for Massive Open Online Courses) research and innovation project, a machine translation approach for online educational content. More specifically, videolectures, assignments, and MOOC forum text is automatically translated from English into eleven European and BRIC languages. Unlike previous approaches to machine translation, the output quality in TraMOOC relies on a multimodal evaluation schema that involves crowdsourcing, error type markup, an error taxonomy for translation model comparison, and implicit evaluation via text mining, i.e. entity recognition and its performance comparison between the source and the translated text, and sentiment analysis on the students' forum posts. Finally, the evaluation output will result in more and better quality in-domain parallel data that will be fed back to the translation engine for higher quality output. The translation service will be incorporated into the Iversity MOOC platform and into the VideoLectures.net digital library portal. Keywords: MOOCs, statistical machine translation, crowdsourcing, CrowdFlower, entity recognition, sentiment analysis 1. Introduction Massive Οpen Online Courses (MOOCs) have been growing in impact and popularity in recent years. According to 2013 statistics1, more than 200 universities around the globe are involved in their creation, with the participation of more than 1,300 instructors, more than 1,200 courses on offer and around 10 million users being actively enrolled. Apart from their significant contribution to lifelong education, MOOCs are viewed as a tool to help identify and fill the gap that exists in the digital skills of workers across Europe. However, the biggest obstacle standing in the way of further growth in online courses is the language barrier, given that the vast majority of such courses are offered in English. Although the need for translating MOOC content has been acknowledged by the majority of course providers2, the solutions provided so far have been fragmentary, human-based, and implemented off-line. TraMOOC 3 constitutes a solution to online course content translation that aims at eleven target languages, is automatic –i.e. it is based on statistical machine translation (SMT) techniques– and is therefore easily extendable to other languages, is adaptable to various types of educational content genre, is independent of course domain, and is designed to produce translations online via its integration in the use-case platforms. The remainder of this paper is organized as follows. 1https://www.edsurge.com/n/2013-12-22-moocs-in-2013-breaki ng-down-the-numbers 2http://blog.coursera.org/post/50452652317/coursera-partnering -with-top-global-organizations 3 www.tramooc.eu Section 2 provides a brief overview of the project. Section 3 describes the collected data, and their domain- and genre-specific idiosyncracies, while Section 4 reports some preliminary translation results. The crowdsourcing tasks involved are presented in Section 5, and the multimodal …

Content truncated to 500 words