Study of Personalized Information Filtering System Based on Multi-Agent

Abstract

The growth of Web information resources leads the phenomenon of information overload and resources disorientation. In order to adapt to the existing recommendation system dynamically changing needs of e-commerce sites, provide users with more proactive, intelligent adaptive personalized product recommendation service, and achieve truly reflect the "information to look for" service model, in this paper, a personalized information filtering system based on multi-agent is presented. The filtering system architecture, agent parts, workflow design are given too. The system will recommend a variety of functional modules in the system can be constructed to take the initiative to customer service and a problem solving environment in self-run entity. This system allows the application of the traditional recommendation system is more malleable, autonomy, and more suitable for the needs of users in a dynamic, uncertain environment to use.