TARGET WORD SELECTION IN MACHINE TRANSLATION BASED ON MACHINE LEARNING

Target word selection in English Chinese machine translation is studied using the machine learning approach. The essence of target word selection is multi classifications of a word. Different translations of a word correspond to different classes.Here proposed is an approach using ID3 decision tree classification algorithm when the features are ambiguous and tree structured. Using the semantic categories of words in the context as features, the approach is proved effective when it is applied to target word selection.