Based on semantic web similarity

In this paper, based on a web page to determine similarity is proposed which is based on semantic web similarity judgments. First on the web page of text extraction, is text segmentation, also taking stop words to go out, Similarity judgments for the Web page provides a basis. Then based on Vector Space Model using TF-IDF method similar to the Chinese website text judgment, In the process of judging, We are adopting a dictionary Words of Synonyms thesaurus, the unity of the key words on the page synonymous, And replacement of pages without the text a synonym for comparison, Experimental results show, The text on the page replacing the unity of synonyms significantly improved the accuracy of the web page similarity judgments.