Web Usage Mining Based on Fuzzy Clustering

The World Wide Web has become the default knowledge resource for many areas of endeavor, organizations need to understand their customers´ behavior, preferences, and future need, but when users browsing the Web site, many factors affect their interesting, and different factor has different degree of influence, the more factors we consider, the more precisely can mirror the user´s interest. This paper describes our attempt to cluster similar Web user, by considering two factors that the page-click number and Web browsing time, which is stored in the Web log, and the different degree of influence of the two factors. The technique proposed in this paper can help Web site organizations to recommend Web pages, improve Web structure, so that can attract more customers, and increase customers´ loyalty.