Research And Implementation Of Agent-based Intelligent E-commerce System

With the development of Internet and information technology, e-commerce has gradually been accepted and deepened in people’s lives. However, the information in the e-commerce is complex, users have to select their own interest from the proliferation of online information then compare and order. Those actions consume customers’effort and increase the network traffic. Therefore we can say the traditional C/S mode can not meet the needs of the development of electronic commerce. The establishment of efficient, fast, intelligent e-commerce has become an inevitable trend of development of electronic commerce.Intelligent Agent has a unique technological advantage of the intelligence, mobility, autonomy. This makes it suitable for application of e-commerce. The combination of e-commerce and Agent gives full play to the characteristics of the Agent. Agent can complete the merchandise selection and negotiation tasks in the transaction process and provide a good solution to the problems that have been arisen in the traditional e-commerce development process.Based on the concept, communication and development platform of Agent, the author proposed bilateral multi-project negotiation model based on genetic algorithms in the negotiation process. By the establishment of a reasonable utility function and learning from each other’s preferences, this model can make the interests to both sides to achieve a relative maximum. This model makes e-commerce transaction intelligent. The model solves the problem of the traditional e-commerce in communication and consultation by the use of Agent method.This system uses MyEclipse as a platform to complete the development and uses Model2-based web development model to complete the intelligent e-commerce system. We use the open source Tomcat as web server, MySQL as the database, use JDBC as the connection to the database. The system uses JADEGateway to complete the integration of JSP/servlet and web application.At first, this paper describes the system development background, propose bilateral multi-project negotiation model based on genetic algorithms then conduct the needs analysis and the overall design. On this basis, we determine the functions of system’s various modules to be achieved, and complete entire coding to achieve intelligent Agent-based e-commerce system. Finally, through the experiment to simulate the negotiation model, we verify the effectiveness and feasibility of negotiation strategies. Results of this thesis enriche the theory of Agent technology and e-commerce content, and propose innovative research and sufficient scientific evidence for the Agent technology in e-commerce application. We can say that this result has a practical value.