ADAPT Continues to Transform Global Content in Preparation for this year’s MT Summit

Posted: 13/07/17

One major area of research carried out in the ADAPT Centre explores the area of Transforming Global Content. Led by Prof Andy Way, ADAPT researchers have successfully published 9 papers and hosted 1 workshop in advance of this year’s Machine Translation (MT) Summit. The completed projects of this  in the areas of MT Modelling, MT Training, Data Scarcity, and Human Factors are crucial in extending research in this global field, and facilitate major advances in terms of human translation and business impact.

Speaking about the accomplishments of the team of expert researchers, Prof Andy Way stated that: “The ADAPT Machine Translation team at DCU are renowned as one of the leading groups on the planet. Our industry engagement is second to none, so a very fruitful area for publications at the moment is our targeted projects with industry, which have led to a large number of publications in leading journals and top-level conferences. The Research Fellows and PostDocs, and the PhD students attracted to DCU are all extremely talented, and we encourage our students to try to publish even in the first year of their studies. This shows them that their work is relevant, acknowledged as top-notch by their peers in the field, and gives them an early opportunity to present their work at leading conferences, where they get to meet other big names in the field, which offers them all the extra encouragement they need to accelerate their rate of progress, and produce more, and better quality, papers in the future. Clearly we’re lucky to have the core ADAPT grant, but the MT team have brought in €2 Million in additional funding (via H2020 and other Irish Government funding) in just two and a half years, which is more than the core grant intended for the full six years, so we’re showing a net profit already!”

Those publications are listed as the following:

• Workshop on Machine Translation for Academic Writing (MT4AW), co-organised by Carla Parra Escartín, Sharon O’Brien (both ADAPT/DCU), Michel Simard (National Research Council, Canada), Marie-Josée Goblet (Université du Québec en Outaouais)
• Panel on Machine Translation and Artificial Intelligence: Probing Near- and Medium-Term Impacts, co-organised by Sharon O’Brien (ADAPT/DCU) and Michel Simard (National Research Council, Canada)
• Machine Translation as an Academic Writing Aid for Medical Practitioners, Carla Parra Escartín, Sharon O’Brien (both ADAPT/DCU), Michel Simard (National Research Council, Canada), Marie-Josée Goblet (Université du Québec en Outaouais)
• The Interact Project and Crisis MT, Sharon O’Brien, Chao-Hong Liu, Andy Way (all ADAPT, DCU) João Graça, André Martins, Helena Moniz, (all Unbabel), Ellie Kemp and Rebecca Petras (both Translators without Borders)
• The Impact of MT Quality Estimation on Post-Editing Effort, Carlos Teixeira and Sharon O’Brien (both ADAPT, DCU)
• Usability of web-based MT post-editing environments for screen reader users, Silvia Rodriguez Vázquez (University of Geneva, formerly DCU) & Sharon O’Brien (ADAPT, DCU)
• Neural Pre-Translation for Hybrid Machine Translation, Jinhua Du and Andy Way (both ADAPT, DCU)
• Elastic-substitution decoding for Hierarchical SMT: efficiency, richer search and double labels, Gideon Maillette de Buy Wenniger (ADAPT, DCU), Khalil Simaan (University of Amsterdam) and Andy Way (ADAPT, DCU)
• Temporality after Translation: A Case Study on Hindi Texts, Sabyasachi Kamila, Sukanta Sen (both IIT Patna), Mohammed Hasanuzzaman, Andy Way (both ADAPT, DCU), Asif Ekbal, and Pushpak Bhattacharyya (both IIT Patna)
• A Comparative Quality Evaluation of PBSMT and NMT using Professional Translators, Sheila Castilho, Joss Moorkens, Federico Gaspari (all ADAPT, DCU), Rico Sennrich (University of Edinburgh), Vilelmini Sosoni (Ionian University, Greece), Yota Georgakopoulou (Deluxe), Pintu Lohar, Andy Way (both ADAPT, DCU), Antonio Valerio Miceli Barone (University of Edinburgh) and Maria Gialama (Deluxe)

The 16th MT Summit, organised by the Asia-Pacific Association for Machine Translation (AAMT), will be held at Nagoya University, Japan, from September 18th through 22nd, 2017. This international conference annually brings people together from both academic and commercial backgrounds who work to develop MT, as well as users of MT systems. The conference overviews state-of-the-art technologies in MT, as well as its major contemporary trends and practical applications.

Learn more about MT Summit 2017 at the following link: http://aamt.info/app-def/S-102/mtsummit/2017/