Agent-Based General Environment with Market Mechanism

The rapid development of e-commerce calls for an immediate need of an autonomous and robust agent system with market intelligence that can offer various types of services to both buyers and sellers. In this paper, Agent-Based General Environment with Market Mechanism (ABGEM) is proposed, with intention to provide a feasible solution to it. The key elements of ABGEM include an agent interaction model inspired from the idea of “Forum”, a set of standard agent types with market mechanism, and the related rules that instruct agents´ activities. To test the feasibility of the proposed solution, an application is developed on ABGEM, simulating two market scenarios. In the experiments, special interest is placed on agent learning of pricing strategies. The result generates valuable findings, which demonstrate the great potential of ABGEM in practical use.