Research On Role-based Negotiation Model In Intelligence E-Commerce

Agent-Oriented Software Engineering (AOSE) and Multi-Agent System (MAS) are being paid more and more attention in software engineer field, Agent-Oriented software development is researched and applied widely too. Its sociality gradually brings to public attention in the research of Multi-Agent System. Individual carries on the action and interaction by taking some social role in human society, comparing with human society, Multi-Agent System is regarded as an organization with sociality. The thesis proposes Role-based Multi-Agent System.Agent technology makes the E-Commerce more intelligent. Multi-Agent negotiation model is applied in the stage of E-Commerce negotiation, which is one part of the Intelligent E-Commerce, dispalying the intelligent character of Agent. By Agent taking the role of negotiation to carry on the intelligent negotiation with simulating the human behavior, the efficiency of negotiation is greatly enhanced in E-Commerce.In this thesis, Multi-Agent System is regarded as the organization with sociality, the tasks and the goals of system are decomposed to roles. Agent completes tasks and realizes goals by taking roles. According to the analysis and the research on Gaia method and BDI model, the role concept based on Gaia is enriched and joined in BDI model, the Role-based Multi-Agent interaction model is proposed. Simultaneously, according to the research on Multi-Agent negotiation model, the Bilateral-multi-issue Negotiation Model (BNM) of E-Commerce is proposed. It defines, analyzes and designs the negotiation model, the negotiation protocol, the negotiation algorithm, etc. Finally, the thesis designs and realizes a demo system of Intelligent E-Commerce negotiation model, using the theory of the Role-based Multi-Agent interaction model that is proposed, implementing and confirming the theory.