Mechanism Design of Online Multi-Attribute Reverse Auction

To solve the problems in government procurement or supply chain management, this paper proposes a multi-attribute online reverse auction mechanism based on multi-attribute decision-making methods, principal-agent theory and online auction technologies. Incentive compatibility constraint is imposed to encourage real weight-setting by bidders. Bargain model and evolutionary learning game are adopted to explain the convergence of the multi-round auction. A multi-attribute online reverse auction mechanism is designed. To implement the method, the architecture based on multi-agent system has been introduced, which includes four agent types. Results show that the mechanism design is incentive compatibility, and it can reduce the moral hazard of principal and risk of winner´s curse. Multi-agent system can help to realize intelligent rule-setting and propositional bidder strategies in online procurement auction. In addition, it is pointed that empirical study on the auction mechanism´s efficiency and Internet implement are further needed.