The Research Of B2C E-commerce Negotiation Model Based On Mobile Agent

With the continuous development of computer technology, especially the extensiveapplication of distributed computing and artificial intelligence, the computer network isgradually infiltrated into every aspect of people’s daily lives. At the same time, e-commercehas grown and online shopping has no longer a strange topic. According to the survey data，the size of China’s online shopping users will reach187million. In the online shoppingmarket, B2C market is growing rapidly, and will become the main driving force of the onlineshopping industry. Negotiation is a complex process which involves many factors. Traditionale-commerce businesses consulted with each other through telephone or C/S mode. By thisway will cause the waste of human, financial and material resources inevitably. Mobile Agenthas the characteristics of intelligence, autonomy and mobility and so on. The development ofthis technology will solve this problem. The application of Mobile Agent technology to thee-commerce environment will become a new direction of electronic commerce development,especially in the field of e-commerce automatic negotiationBased on the learning theory and technology of electronic negotiation, and simulatingtraditional negotiation process, this article established a B2C e-commerce negotiation modelbased on Mobile Agent. The model is described as a nine-tuple. It defined the elementsrespectively which may be involved in the consultation process. In order to achieve amultilateral consultation, multithreading idea is introduced into the negotiation model, thatmakes all of the parties who meet the initial conditions can take part in the consultation. Inthe negotiation process, the agents made decisions by the utility evaluation of the proposalsboth from their own and other parties. In order to achieve the maximize utility, a dynamicconcession strategy based on opponents attitude learning is proposed. By learning the track ofthe proposals, Agents can understand each other’s preferences, attitude and urgency. And thengive their own counter-proposals based on the learning. The negotiation model can support acontinuous and dynamic process. All of the parties who participated in the negotiation will ina competitive environment, because their pricing and concession strategies affect the outcomeof the consultations. And this way help the negotiation parties to achieve a "win-win" situation.In order to verify the negotiation model, we designed a B2C e-commerce negotiationmodel system based on Mobile Agent. The system was written by Java language and usedAglet system as a support platform. Finally, we built a test environment of the model system.Through the functional test and the incremental test, we verified that the system is correct andhave operability and scalability.Finally, the main conclusions are summarized. And because of the limited time andresources, this article also has some deficiencies.