Improving a Lexicalized Hierarchical Reordering Model Using Maximum Entropy
Vinh Van Nguyen, Akira Shimazu, Le-Minh Nguyen, Thai Phuong Nguyen · 2009
In this paper, we present a reordering model based on Maximum Entropy. This model is extended from a hierarchical reordering model with PBSMT (Galley and Manning, 2008), which integrates syntactic information directly in decoder as features of MaxEnt model. The advantages of this model are (1) maintaining the strength of phrase based ap-proach with a hierarchical reordering model, (2) many kinds of linguistic information inte-grated in PBSMT as arbitrary features of Max-Entropy model. The experiment results with English-Vietnamese pair showed that our ap-proach achieves improvements over the sys-tem which use a lexical hierarchical reorder-ing model (Galley and Manning, 2008). 1