User Classification Algorithm Based on Web Browsing Behavior

Current Internet is ubiquitous and the number of web users increases quickly, producing abundant information. Therefore, it is necessary to classify the users and find the excellent users. However, the amount of user data is enormous and the un-certainty exists in user data makes this work difficult. In user classification analysis, the key problems to be solved include how to measure classification conducts and designing classification algorithm according to browsing behavior. User behavior data contains noisy information, it is important to reduce the noise data in user information. In this paper, we propose a rough set based algorithm of data reduction and a fuzzy-SVM based algorithm on user classification and evaluation. The algorithm can decrease the redundant attributes and give an adequate evaluation on user class.