Research Of User Behavior Analysis System Based On Interest Classification

With the rapid development of the Internet and the popularization of the computer network, the scale of network users extended gradually, and network user's behavior becomes more and more complicated. Meantime, the profitable websites need to offer more satisfactory service through analyzing users' behaviors and hobbies in order to obtain maximum profits; and the nonprofit websites such as government, scientific research institution, need to analyze users to build a scientific Decision Support System. For these reasons, the research and analysis on network users' behaviors becomes more and more important.Web log mining discovers user profiles while accessing the web pages by analyzing Web log generated by Web server. Though analyzing these user profiles, we can discover some regulations, which can help us find out more potential users, improve quality of service to the ultimate users and improve the performance of the Web server.This paper focuses on analysis of user behavior and interest in data mining classification techniques, network users and network proposed the concept of user behavior, and on this basis, reading through the extensive literature to study the user behavior analysis, the theory of knowledge, with its in-depth understanding. Study the characteristics of network user behavior and classification, learning the user behavior analysis commonly used in Web Mining and Web crawler technology. Of existing technology to improve web crawlers, web pages achieved using ODP topics classification. Interest in classification is proposed based on user behavior analysis system, the system architecture was designed and completed the implementation of four sub-modules. Including data acquisition module, data preprocessing module, interest mining module, personalized service module. Test run on the system, and the system's applications.