Integrate word sense disambiguation based on parsing tree into machine translation

Word sense disambiguation is important to machine translation. In this paper, a new word sense classifier is proposed, whose discriminative features are extracted from parsing tree. At the same time, the bayesian classifier is adopted to determine correct senses of ambiguous words. Then, the optimized word sense classifier is integrated into a machine translation system. Experimental results show that the accuracy of the proposed word sense classifier is improved and the translation quality of machine translation system increases.