Paper: A Hybrid Word Alignment Model for Phrase-Based Statistical Machine Translation

ACL ID W13-2814
Title A Hybrid Word Alignment Model for Phrase-Based Statistical Machine Translation
Venue Workshop on Hybrid Approaches to Translation
Session
Year 2013
Authors

This paper proposes a hybrid word alignment model for Phrase-Based Statistical Machine translation (PB-SMT). The proposed hybrid alignment model provides most informative alignment links which are offered by both un- supervised and semi-supervised word align- ment models. Two unsupervised word align- ment models (GIZA++ and Berkeley aligner) and a rule based aligner are combined togeth- er. The rule based aligner only aligns named entities (NEs) and chunks. The NEs are aligned through transliteration using a joint source-channel model. Chunks are aligned employing a bootstrapping approach by trans- lating the source chunks into the target lan- guage using a baseline PB-SMT system and subsequently validating the target chunks us- ing a fuzzy matching technique against the target ...