A Research On Concurrent Bi-lateral Automated Negotiation Model For Retail Electronic Market

The development of retail electronic market is now attracting more and more merchants and consumers trading online. The applications of Intelligent Agent technology in e-commerce systems help automate many tasks of business. The negotiation takes a most important part in actual commercial activities, thus, automated negotiation technology has become the key technology in realization of second-generation e-commerce.Many current automated negotiation systems support one-to-one negotiation. One-to-many negotiation has been mostly automated using various kinds of auction mechanisms, which is a kind of competitive negotiation and usually concerns with a single issue of price. Comparing to this, multiple concurrent bi-lateral negotiation is a more flexible and suited model applying in semi-competitive retail electronic market. The central problem addressed in this thesis is the design and engineering of a multiple concurrent bi-lateral negotiation model applying to trading negotiation in retail commerce.The major research work in this thesis can be viewed as follow:1. The thesis analyzes the state of the art of the automated negotiation systems, proposes the multi-agent system-based model of concurrent bi-literal negotiation. There are three kinds of agent in the model: coordinator Agent > buyer agent and seller agent. And there are two kinds of communication among them: the bi-lateral multiple issues negotiation between buyer agent and seller agent, multiple threads coordination between coordinator agent and buyer agent, and we construct a model of each one of these two.2. Based on the study of characters of retail electronic market, the thesis designs a PFCSP-based (Prioritized Fuzzy Constraints Satisfaction Problem) bilateral multiple issues negotiation model, puts forward the formal description and behavior protocols of buyer agent and seller agent, defines the communication interfaces, proposes three negotiation strategies based on fuzzy constraints and corresponding algorithms of buyer agent.3. The thesis constructs multiple threads coordination model, puts forward the formal description and behavior protocols of coordinator agent, defines the communication interfaces, proposes three coordination strategies and corresponding algorithms of coordinator agent.4. Based on the research mentioned above, we designs and implements an experiment system MCBNS (Multiple Concurrent Bi-lateral Negotiation System) using JAVA language on the JADE (Java Agent Developing Environment) platform. Through the experiments, we verify the validity and feasibility of the bi-lateral multiple issues negotiation and multiple threads coordination, and compare varieties of negotiation strategies and coordination strategies.