Agent-Based Relevant Reseatch Of Auto Negotiation

With the development of agent technology, e-commerce based on Agent technology has become the hot area of academic research. Currently the technology support of e-commerce is still weak, especially in the domain of auto negotiation. The advantages of agent such as sociality, automation can effectively adapt the demands of e-commerce, improve the automation of e-commerce.This paper mainly studies the automatic negotiation based on Agent, the two mainly parts of negotiation is negotiation protocol and negotiation strategy. Currently the negotiation protocol is coded in negotiation agent, so in the electronic market environment ,this kind of fixed negotiation protocol in negotiation agents can't satisfied the negotiation demand of heterogeneous agent. If the heterogeneous agent has the same knowledge of negotiation and load negotiation rules dynamically in the negotiation process, they can satisfy the interactions between heterogeneous agents. We use ontology express the knowledge of negotiation. Due to the expressional of traditional ontology language OWL DL, we use the reasoning mechanism framework based on SWRL rules to complement the ontology content.By analyzing the opponent's offers in the negotiation process, a automatic negotiation system based on fuzzy similarity is proposed, which can adjust negotiation concession dynamically according to the given fuzzy control rule. The counter-offer is computed by maximizing the opponent's utility on the promise of remaining the same utility of agents, by doing so,the result is guaranter to be pareto optimize.