Generating English Determiners in Phrase-Based Translation with Synthetic Translation Options
Yulia Tsvetkov, Chris Dyer, Lori S. Levin, Archna Bhatia · 2013
We propose a technique for improving the quality of phrase-based translation systems by creating synthetic translation options—phrasal translations that are generated by auxiliary translation and postediting processes—to augment the default phrase inventory learned from parallel data. We apply our technique to the problem of producing English determiners when translating from Russian and Czech, languages that lack definiteness morphemes. Our approach augments the English side of the phrase table using a classifier to predict where English articles might plausibly be added or removed, and then we decode as usual. Doing so, we obtain significant improvements in quality relative to a standard phrase-based baseline and to a to post-editing complete translations with the classifier. 1