The Study Of Network Information Retrieval Based On Improved Vector Space Model

With the quick development of computing technology and network technique, network information increase rapidly. Consequently, network information retrieval has become the key to obtaining desired information rapidly and efficiently. As one of the important technology in information retrieval, vector space model (VSM) has been widespread application, but it has some shortages. In view of the two important shortages in VSM, this paper points out the improved measure differently. One is the improvement of calculation method about feature weight. Synthetically considering the category difference based on the tfidf calculation method, this paper put forward a new weight calculation method. The other is that semantic expansion is constructed for query vector by the Hownet. As a result, the semantic retrieval model based on concept is realized. We can conclude that the two improvement increase the information retrieval performance efficiently.