The Calculation of User Interest based on Web Logs and Web Page Feature Contents

Knowing user's interest and recommending information what they need exactly is an important task in e-business websites.Author proposes a personalized information recommendation algorithm: basing on web logs,referring to the contents of web page and the time used in browsing pages.Following the principle of "one-session,single-target",this algorithm can get user's interest types of information and merchant scientifically by analyzing the web page which is classified by its feature contents.