The Research On The Automated Trade Theory And Method Based On Agent In The E-Marketplace

The rapid development of Internet has promoted the forming of e-marketplace and changed the traditional business model, and it also has brought great changes and deepened influences in all kinds of aspects of our society. The e-marketplace not only boosts the online trading growth increasingly,but also accelerates speed of goods exchange and greatly facilitates both the buyer and the seller.Now, how to improve the efficiency of e-marketplace and reduce the cost of trade becomes the problem deserving of study. The paper analyses the related problems of e-marketplace theories and solutions , and gives emphasis to discuss the forming process of automated negotiation mechanism about two sides and multi-issues in theory, so we can find out a automated trade matching system with high efficiency and low cost.In theory, the paper constructs an e-marketplace negotiation model with learning ability by applying fuzzy logic, utility function and a preference learning rule and integrating with issue-trading strategies in finding mutually-agreed contracts.It also elaborately discusses the concession strategies of negotiation issues in the qualitative and quantitative ways. In methodology, the paper establishes a negotiation mechanism which can improve both the automated processing flow and the automated trade by using agent technology, and the whole framework of an e-marketplace automated trade is finally built up. At the same time, through the research on the match mechanism of an e-marketplace, it can match buyers and sellers based on the similarity degree of their proposals ,then launches its negotiation in the environment of multi-buyers and sellers, which further enhances the trade efficiency . The research on automated trade theory and method based on Agent will contribute values to the e-marketplace development theoretically and practically.