This paper describes the MIT-LL/AFRL statistical MT system and the improvements that were developed during the IWSLT 2012 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 task. We also applied our existing ASR system to the TED-talk lecture ASR task, and combined our ASR and MT systems for the TED-talk SLT task. We discuss the architecture of the MIT-LL/AFRL MT system, improvements over our 2011 system, and experiments we ran during the IWSLT-2012 evaluation. Specifically, we focus on 1) cross-domain translation using MAP adaptation, 2) cross-entropy filtering of MT training data, and 3) improved Arabic morphology for MT preprocessing
Cite as: Drexler, J., Shen, W., Terry, , Anderson, T., Slyh, R., Ore, B., Hansen, E. (2012) The MIT-LL/AFRL IWSLT-2012 MT system. Proc. International Workshop on Spoken Language Translation (IWSLT 2012), 109-116
@inproceedings{drexler12_iwslt,
author={Jennifer Drexler and Wade Shen and Terry and Tim Anderson and Raymond Slyh and Brian Ore and Eric Hansen},
title={{The MIT-LL/AFRL IWSLT-2012 MT system}},
year=2012,
booktitle={Proc. International Workshop on Spoken Language Translation (IWSLT 2012)},
pages={109--116}
}