Automatic identification of hierarchical relationship between words based on clustering

At present, how to use computer technology to automatically identify the semantic relations between terms, such as equivalence relationship, hierarchical relationship and associative relationship, is a key part of automatically building a thesauri, also a difficulty. In this paper, several common methods to identify the concept hierarchical relationship were described: the identification method based on co-occurrence analysis, the distribution of similarity calculation method and the method based on syntactic pattern matching, proposed a combination method of co-occurrence statistics and distribution of similarity calculation to identify the concept hierarchical relationships between words, and proved by experiments that the method is feasible and effective.