关键词: 多属性拍卖, 数据包络分析, 第二分值拍卖, 信息不对称, 激励相容

Abstract: Multi-attribute auctions incorporate multiple attributes, such as priceand quality, on the basis of traditional price auctions, and have been widely used in many fields.In order to study the situation where the inputs have almost no prior structure and the number of bidders is large enough, this paper expands a single input to multiple inputs based on previous studies on multi-attribute auctions and reasonably divides multiple attributes of items (such as price, completion time, labor quantity, quality) into inputs or outputs. An effective multi-attribute second score auction mechanism is designed based on data envelopment analysis. Compared with other methods, data envelopment analysis combined with the multi-objective programming methods can effectively solve the multi-input and multi-output problems to help the auctioneer maximize his own utility while finding bidders with better overall performance. The mechanism satisfies the constraints of individual rationality and incentive compatibility. Compared with the traditional second score auction mechanism, thenew mechanism can maximize the utility of the auctioneer under the premise of attracting bidders.

Key words: multi-attribute auctions, data envelopment analysis, second score auctions, information asymmetry, incentive compatibility