The Effective Classification of the Chines e Web Pages Based on KNN

In order to improve the efficiency and accuracy of classifying the Chinese web pages and help users to locate pages of their interest quickly, this paper presents an efficient feature selection method. We assign weights to different HTML tags and compute the final weight of each word occurred in the document, and then select the representative feature words to describe the document. The method combing the KNN classification algorithm can classify the Chinese web pages effectively. Experimental results demonstrate that the method can reduce the dimension of space and improve precision and recall obviously.