Research of Word Sense Disambiguation Based on Mining Association Rule

Word sense disambiguation (WSD) is always an important and difficult problem that requires to be solved in natural language processing. This paper presents a new WSD method which is based on the mining association rules. The method can mine the association rules between the sense of the ambiguous word and its context, to construct an association rules-based database. At last the sense of the ambiguous word is marked by choosing the sense which the most association rules define. The experiment result shows that the method has high precision.