A Negotiation Model and Its Simulation Based on Q-Learning Algorithm

With the applications of e-commerce and the development of network technology such as the Semantic Web and Ontology, the use of artificial intelligence to replace the traditional approach of electronic trading has become a hot spot within related research. On the basis of Agent technology, this article focuses on Automatic Negotiation and analyzes the framework of Negotiation Model based on Multi-Agent tech. Based on the model, this article uses the Q-learning algorithm within Artificial Intelligence and multi-agent system (MAS) simulation platform -Swarm, investigates, brings about the negotiation agent with ability of autonomous learning and verified the feasibility of this model through simulation. The conclusion shows that the model has a sound convergence and rationality.