Better Filtration and Augmentation for Hierarchical Phrase-Based Translation Rules

This paper presents a novel filtration criteria to restrict the rule extraction for the hierarchical phrase-based translation model, where a bilingual but relaxed well-formed dependency restriction is used to filter out the bad rules. Furthermore, a new feature which describes the regularity that the source/target dependency edge triggers the target/source word is also proposed. Experimental results show that, the new criteria filters out about 40% rules while with translation performance improvement, and the new feature brings another improvement to the baseline system, especially on larger corpus.

• Type: Journal
• Year: 2010
• Where: ACL
• Authors: Zhiyang Wang, Yajuan Lv, Qun Liu, Young-Sook Hwang
• Keywords: ACL 2010, Computational Linguistics, Hierarchical Phrase-based Translation, Translation Performance Improvement, Well-formed Dependency Restriction