What End Users Really Think of Raw and Post-edited MT
Martina Bellodi & Sabrina Girletti
Tuesday, 26 March, 2019
Swiss Post writes process manuals and instructions in German, and then translates them into Italian and French for the Ticino and Romandie regions. Translation is currently a fully human process. Swiss Post Language Services has been testing neural machine translation (NMT) and plans to introduce it in their standard translation process in order to improve turnaround times. So far, the tests have focused on both the usability of raw NMT output for post-editing by in-house linguists and the actual productivity gains in the translation workflow thanks to MT post-editing. In a further step we will conduct a study with Swiss Post’s French- and Italian-speaking employees to assess end users’ acceptance of raw vs. post-edited NMT output. The purpose of the study is to shed light on some neglected aspects of NMT research, i.e. the perception of quality among end users of raw machine translation, the impact of post-editing on the perception of NMT quality, and the risks associated with the use of raw NMT in a business setting.
In this presentation the goals and results of the study will be explained and discussed.
Martina Bellodi graduated from the University of Bologna in 2003 and began her career as a freelance translator. In 2009 she started working as an in-house translator at Swiss Post Language Services. She was promoted to Head Translator in 2011 and to Deputy Head of Language Services in 2012. Since 2014 Martina has been in charge of Language Services’ operational and strategic management. She will graduate with an EMBA at the beginning of 2019 and has appeared as a keynote speaker at several industry conferences (tcworld Stuttgart, LQA Symposium Zurich, XTM Live Amsterdam).
Sabrina Girletti is a PhD student and teaching assistant at the Translation Technology Department of the Faculty of Translation and Interpreting (FTI), where she contributes to postgraduate courses in machine translation and localisation. Her research interests include post-editing approaches and human factors in machine translation. She is currently involved in a project testing the implementation of machine translation at Swiss Post. Sabrina holds a master’s degree in Translation, with specialisation in Translation Technologies, from the University of Geneva, and a bachelor’s degree in Linguistic and cultural mediation from the University of Naples "L’Orientale".
Most machine translation output is fine for basic Western LSP needs, as they usually translate within European...
Machine translation is great when it works, but it fails often and in unpredictable ways. What causes bad machine...