Paper: Assessing the Discourse Factors that Influence the Quality of Machine Translation

ACL ID P14-2047
Title Assessing the Discourse Factors that Influence the Quality of Machine Translation
Venue Annual Meeting of the Association of Computational Linguistics
Session Main Conference
Year 2014
Authors

We present a study of aspects of discourse structure ? specifically discourse devices used to organize information in a sen- tence ? that significantly impact the qual- ity of machine translation. Our analysis is based on manual evaluations of trans- lations of news from Chinese and Ara- bic to English. We find that there is a particularly strong mismatch in the no- tion of what constitutes a sentence in Chi- nese and English, which occurs often and is associated with significant degradation in translation quality. Also related to lower translation quality is the need to em- ploy multiple explicit discourse connec- tives (because, but, etc.), as well as the presence of ambiguous discourse connec- tives in the English translation. Further- more, the mismatches between discourse expressions ...