Experiments with POS-based restructuring and alignment-based reordering for statistical machine translation
Shuo Li, Derek F. Wong, Lidia Sam Chao · Workshop on Hybrid Approaches to Translation · 2013
This paper presents the methods which are based on the part-of-speech (POS) and auto alignment information to improve the quality of machine translation result and the word alignment. We utilize different types of POS tag to restructure source sentences and use an alignment-based reordering method to improve the alignment. After applying the reordering method, we use two phrase tables in the decoding part to keep the translation performance. Our experiments on Korean-Chinese show that our methods can improve the alignment and translation results. Since the proposed approach reduces the size of the phrase table, multi-tables are considered. The combination of all these methods together would get the best translation result.