Research on the Model of Personalized Recommendation System Based on Multi Agent in E-Commerce

Abstract

There are some lacks of intelligence, self-adaptability, initiative and processing power limitations in the traditional recommendation system. Using the multi-agent technology and the web log mining technology, this paper converts the function modules of traditional personalized recommendation system into an agent. This paper proposes an architecture model based on multi-agent e-commerce personalized recommendation system (MAPRS), and discusses the function of each component of the model and the system's running processes.