Research on electronic commerce automated negotiation in multi-agent system based on reinforcement learning

• Published in: 2009 International Conference on Machine Learning and Cybernetics Vol. 3; pp. 1419 - 1423
• Main Author: Jin-Gang Cao
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.07.2009
• Subjects: Automated negotiation; Business; Cybernetics; Electronic commerce; Intelligent agent; Learning systems; Machine learning; MAS; Multiagent systems; Paper technology; Protocols; Reinforcement learning; Utility theory
• ISBN: 9781424437023; 1424437024
• ISSN: 2160-133X
• DOI: 10.1109/ICMLC.2009.5212335

In order to improve the efficiency and intelligence of negotiation, the paper applies the technology about agent and the mechanism of reinforcement learning to electronic commerce negotiation. Through presenting negotiation protocol and analyzing negotiation flow based on multi-attribute utility theory, the paper builds an open and dynamic automated negotiation model, and imports Q-learning into the negotiation to quicken the process of negotiation. Compared with no learning mechanism in negotiation, the negotiation efficiency of the model has been improved and the negotiation results are acceptable.