Design And Implementation Of Multi-agent-based E-commerce Personalized Recommendation System

In today's fast-growing process of e-commerce,the design and development of intelligent personalized recommendation system is of great value in improving the operating performance.Aiming at the existing recommendation system lacking of initiative,intelligence,etc,the business of intelligent e-commerce personalized recommendation system,the overall structure,functions and implementation of the system are researched by using the design idea of multi-Agent.The paper's concrete work is as follows:Firstly,a e-commerce personalized recommendation model based on muti-agent technology is put forward by the paper.By adopting muti-agent technology and combining with web log,the components of functional modules of traditional personalized recommendation system is constructed of intelligent agent.Meanwhile,a e-commerce personalized recommendation overall structure model—MAPRS(Multi-Agent based Personalized Recommendation System, MAPRS)is presented.The model is composed of eight functionally independent agents,implement personalized and initiative recommendation function by their cooperation,and overcome the limitation of traditional recommendation system.Secondly,the function and structure of each Agent in the MAPRA model above mentioned and the work flow of the whole system are designed.According to the practical needs of recommended e-commerce system,the function and structure of monitor agent,web log analysis agent,data conversion agent,data mining agent, preference analysis agent,interface management agent and the recommendation engine agent are designed and the overall operational process of MAPRA model is discussed.Thirdly,a e-commerce personalized recommendation system- PR Market is built based on the lightweight J2EE frame and above mentioned model.In order to verify the model presented by paper and according to the above MAPAS model,a prototype system of personalized e-commerce recommendation system is designed and implemented,called the PR Market system,by using lightweight J2EE Software arehitecture,J2EE runtime environment,Eclipse5.5 development platform and Oracle 9i database.The study result of this paper can provide certain reference for the e-commerce enterprise to design and develop e-commerce personalized recommended system.