Using unlabeled dependency parsing for pre-reordering for Chinese-to-Japanese statistical machine translation
Dan Han, Pascual Martínez-Gómez, Yusuke Miyao, Katsuhito Sudoh, Masaaki Nagata · 2013
Chinese and Japanese have a different sentence structure. Reordering methods are effective, but need reliable parsers to extract the syntactic structure of the source sentences. However, Chinese has a loose word order, and Chinese parsers that extract the phrase structure do not perform well. We propose a framework where only POS tags and unlabeled dependency parse trees are necessary, and linguistic knowledge on structural difference can be encoded in the form of reordering rules. We show significant improvements in translation quality of sentences from news domain, when compared to state-of-the-art reordering methods. 1