Research On Text Mining Based Web Information Retrieval

It makes search engine's problems such as low veracity more and more serious that the information in the Internet is getting richer and richer. This paper does research on how to resolve the problem of low veracity in Web information retrieval with text mining.Web text mining, especially Web text clustering, Web information retrieval principles and Web text models are introduced in detail. Vector Space Model is improved according to the characteristics of the Web text model. Fuzzy C-Means clustering and Self-organized Map Neural Network are combined and a clustering algorithm of Improved Fuzzy Self-organized Map Neural Network is proposed.Finally implement a clustering search engine experimental system with the Improved Fuzzy Self-organized Map Neural Network. The improvement of veracity by above model and algorithm is verified.