An Efficient Web Page Recommendation Based on Preference Footprint to Browsed Pages

This paper proposes a new scheme for web page recommendation which reflects the preference of each user to the recommended pages in an efficient and effective manner. The basic idea of the scheme is to combine the notion of preference footprint to browsed pages with the collaborative filtering. More concretely, we introduce the notion of "tags" similar to conventional SBS (Social Bookmark Service), and attach all tags associated with a user to a page when it is browsed by him. We implemented a prototype of the proposed scheme, and conducted preliminary experiments to evaluate the performance of the scheme. The result of experiments indicates that it takes less than 0.5 sec to reorder a list of 500 URLs received from a search engine according to the preference of users.