Web Page Categorization Based on k-NN and SVM Hybrid Pattern Recognition Algorithm

Traditional information retrieval (IR) method use keywords matching to filter the documents, but usually retrieves unrelated Web pages. In order to effectively classify Web pages, we present a Web page categorization algorithm, named WebPSC (Web page similarity categorization). This algorithm uses latent semantic indexing (LSI) to model Web pages and implement categorization based on hybrid pattern recognition algorithm of the k-NN and support vector machine (SVM). As an implementation of WebPSC, an intelligent agent system to acquire user interest and help user retrieving Web pages is presented. Empirical results of using this algorithm indicate our method can reach high levels of accuracy in Web page classification.