Abstract: To improve the performance of webpages classification system,and overcome a large number of false, erroneous information filled in the webpages affect the traditional classification algorithms,this paper presents a web page classification algorithm based on link information.Based on the K Nearest Neighbor method,the webpages are classified by the links among webpages.In this paper,the webpage currently classified is presented by the link information of vector space,and find K webpages with the highest similarity to it in the training set.then it is classified to the proper category.We compare the method to traditional classification algorithms through experiments,and the results show that it s more effective.