International Workshop on Spoken Language Translation (IWSLT) 2006
Keihanna Science City, Kyoto, Japan
We present the CMU-UKA Syntax Augmented Machine Translation System that was used in the IWSLT-06 evaluation campaign. We participated in the C-Star data track using only the Full BTEC corpus, for Chinese-English translation, focusing on transcript translation. We applied techniques that produce true-cased, punctuated translations from non-punctuated Chinese transcripts, generating translations which score higher against the Official metric than against the lower-cased, punctuation removed metric. Our results demonstrate the impact of syntax and hierarchy based models for speech transcript translation.
Full Paper Presentation
Bibliographic reference. Zollmann, Andreas / Venugopal, Ashish / Vogel, Stephan / Waibel, Alex (2006): "The CMU-UKA syntax augmented machine translation system for IWSLT-06", In IWSLT-2006, 138-144.