Market Confidence and Reputation Feedback System in Online Transactions - Based on Multi-agent Computational Experiment

In order to explore the role of the market confidence and reputation feedback systems in online transactions, a computational experiment model based on the multi-agent is used in this paper. By introducing experience weighted attraction learning algorithm which endows the agent with the bounded rationality, the model can represent the revolution trends in online market and analyze the effect of the initial market confidence and the reputation feedback systems comparatively. Empirical results show that the reputation feedback systems may accelerate the decline of the online market with the lack of initial market confidence. The model can explain why the online transactions bloom only after introducing the escrow in China, therefore it can help improve management.