Deep Chinese Word Sense Disambiguation Method Based on Sequence to Sequence

In the field of natural language processing, word sense disambiguation plays an important role. The word sense disambiguation method based on traditional machine learning is not high in accuracy, and it is necessary to extract text features manually; the method based on deep learning has not been applied to the case where there are many ambiguous meanings. For the characteristics of Chinese text, the deep Chinese word sense disambiguation method based on sequence to sequence is proposed in this paper. The input is a word context sequence, and the output is a word meaning sequence, which is applicable to all word meaning ambiguity cases. Finally, the method is compared with other seven methods. Test with the data set in the SemEval-2007 Task #5 task. The results show that the test accuracy of the disambiguation is improved by 11.48% compared with the method with the highest accuracy among the seven methods.