Study On Theories And Methods Of Multimode Intelligent Negotiation Support System

As an indispensable part of social life, negotiation is an important form of human association. The history of human civilization is also the history of human negotiation. Due to the differences between objective and subjective aspects, individuals or organizations have the conflict between goals or interests. Negotiation, as an important form of resolving the dispute or conflict, has been widely applied to social life.Generally, negotiation is the face-to-face negotiation held by all sides in the same place. They exchange their ideas, hold conferring and sign agreements, etc. In recent years, with the fast development of computer network technology, a new information age comes and has promoted the popularization of Internet and the rapid development of B to B and B to C Electronic Commerce. As Electronic Commerce becomes more and more complicated and the two sides of trade need real-time and interactive communication to confirm the trade or the contract, the remote negotiation through the network should be required. Since 1980s, Negotiation Support System (NSS), which is regarded as a branch of Group Decision Support System, is focused on. With the development of relevant theory, experiment systems achieve a lot eventually. These systems provide the mechanism based on process support and solution support for the negotiation that is composed of human attending, communicating and signing contracts. Commonly, artificial negotiation can easily describe the complicated multi-objective issues and can flexibly control the negotiation process meanwhile, but it takes up a lot of resources and time, so it is necessary to introduce automatic negotiation to improve conference efficiency, decrease operation complexity and negotiation expenses. On the basis of different negotiation methods, this thesis proposed and established Multimode Intelligent Negotiation Support System (MINSS), which considers both complexity and variability of the conventional negotiation activity and the fast and simple practical demands of the new generation Electronic Commerce mode, and provides remote negotiation and conference with a flexible, high-efficient and economic support platform. The major research content of this thesis are described as follows:1. After the present NSS and relevant technologies were analyzed, the research work and achievements of NSS were summarized and the disadvantages were pointed out.2. Based on the commercial negotiation theory and through analyzing the current situations of NSS at present, NSS architecture adopting negotiation agent was proposed and the concept of MINSS was presented, which integrates artificial negotiation mode, semi-automatic negotiation mode and full-automatic negotiation mode and forms a research thought and goal of new type NSS and has a beneficial trial to improve the performance of current NSS.3. The system flow of multimode intelligent negotiation was analyzed and the framework structure of each mode was constructed. Based on machine learning, DSS theory and agent technology, the function of the three modes was analyzed and regarded as a organic whole, and a design thought of intelligent NSS was presented, which uses artificial negotiation to instruct and train semi-automatic negotiation and replaces the former with the latter.4. With the induction learning method, the mapping from database to the knowledge base was set up and a new data structure was defined to represent the knowledge. For artificial negotiation subsystem,the representing and saving methods of historical cases were focused on. For learning subsystem, 1D3 algorithm was introduced to the negotiation field in order to develop the new thought of machine learning.5. In the semi-automatic mode which substitutes the negotiation persons on one side with the agent, bargaining plays a decisive role to resolve the problems. The bargaining strategy including Method-1, Method-2 and Method-3 was designed and the interactive communication of negotiation between agent and physical human was realized.6. An initial design of the full-automatic mode was proposed, in which the two sides of negotiation are agents. Based on the mathematical models, the multi-objective negotiation model and the genetic algorithm were applied to work out the Pareto optimum solution.7. M1NSS which was realized by programming, was developed on the mobile agent platform— Aglets, and resolved the problems of database generating and mobile agent producing. Some application technologies of MINSS were explained, and architecture framework of MINSS and a part of man-computer interface were showed when the system was operating. The ideas of multimode intelligent negotiation were realized, initially.In conclusion, this thesis presents the concept of MINSS for the first time in the relevant field of negotiation. Through constantly expanding, the multimode negotiation theory was formed, which was composed of negotiation agent, machine learning and Pareto negotiation solution model. And the bargaining strategy was designed, which was used in auto-negotiation. These works play a positive role to improve the current theoretical methods of NSS and the negotiation technologies of Electronic Commerce.