Page Interest Estimation Based on the User's Browsing Behavior

A page interest estimation method based on analyzing the users' browsing behaviors is designed in this paper. Different from existing methods, present system avoids users' feedback to the browser and doesn't collect the information which may produce privacy issues, e.g. users' browsing history, bookmarks, and so on. Because the information recorded by server in access logs will be utilized to evaluate the page interest, a referer-based data preprocessing method is carried out to improve the reliability of the access data and extract the necessary information for interest estimation. An interest estimation model which mainly considers the reference length, the size, the visiting time and the visiting times of each accessed page is then used to determine the page interest. The proposed method is tested by analyzing the practical web access logs for validation. The page interest estimation can be efficiently accomplished and the data set is well prepared for further machine learning algorithms to construct user profiles.