A Multi-agent, Multi-object and Multi-attribute Intelligent Negotiation Model

Automated negotiation among software agents is becoming increasingly important in e-business. Currently, most of these multi-agent negotiation models are focused simply on mono-object negotiations. But in fact, most practical business negotiations are based on multiple objects. In order to improve the intelligent negotiation model to meet the practical requirements, this paper proposes an integrated multi-agent, multi-object and multi-attribute intelligent negotiation model named M3INM. M3INM introduces a scalable multi-agent negotiation framework and integrates various negotiation rules into the model. In order to meet the balance among multiple attributes, the model introduces the idea of simulated annealing (SA) iterative algorithm. This deliberate SA algorithm adjusts dynamically the correlativity among attributes to obtain the optimized negotiation solution. Besides, the relevant negotiation mathematics model is also discussed and simulation results are presented and analyzed in this paper.