The MIT-LL/AFRL IWSLT-2011 MT System
MASSACHUSETTS INST OF TECH LEXINGTON LINCOLN LAB
Pagination or Media Count:
This paper describes the MIT-LLAFRL statistical MT system and the improvements that were developed during the IWSLT 2011 evaluation campaign. As part of these efforts, we experimented with a number of extensions to the standard phrase-based model that improve performance on the Arabic to English and English to French TED-talk translation tasks. We also applied our existing ASR system to the TED-talk lecture ASR task. We discuss the architecture of the MIT-LLAFRL MT system, improvements over our 2010 system, and experiments we ran during the IWSLT-2011 evaluation. Specifically, we focus on 1 speech recognition for lecture-like data, 2 cross-domain translation using MAP adaptation, and 3 improved Arabic morphology for MT preprocessing.
- Voice Communications