Abstract：With the size of WWW growing bigger, its structure is more and more complex. Web visitors with personalized information demands find it very difficult to locate information with ease. In this paper, we designed an intelligent web page recommender system, which include three sub-systems of data preprocessing, clustering, and web page recommendation. The proposed system can divide visitors into subgroups automatically, thus providing new users with online web page recommendation. The formula for computing the distance between visitor paths is proposed, followed by the study of the structure of clustering system. The simulation results based on web log of Microsoft website proved the validity of the proposed method.