An Improved Word Sense Disambiguation Method

Abstract

Word sense disambiguation (WSD) is a critical and difficult issue in natural language processing(NLP), and WSD is great significance in large area of research areas of NLP without doubt. The method here based on match word corpus and the calculation of similarity and relevance. In addition the method proposed in this paper believes that all sentences in the same paper are correlated, especially the sentences with the same ambiguous words. So the same ambiguous word has the same meaning, while the sentences including them own the highest similarity. The experiments show that the proposed WSD method can obtain good results.