A tree-to-string phrase-based model for statistical machine translation

A tree-to-string phrase-based method which is based on synchronous CFGs. This method has two important properties: syntactic transformation is used in the decoding phase and phrases are the basic unit of translation. We carried out experiments with two language pairs English-Japanese and English-Vietnamese. Our system achieved significant improvements over Pharaoh. We also analyzed the dependence of translation quality on the level of syntactic analysis (shallow or deep).

Auto71

https://www.aclweb.org/anthology/W08-2119.pdf