基于数据包络分析的混合多属性第二分值拍卖机制

Hybrid Multi-attribute Second Score Auction Mechanism Based on Data Envelopment Analysis

多属性拍卖在传统价格拍卖的基础上纳入了价格和质量多个属性,已广泛应用于许多领域。本文针对输入几乎没有先验结构且投标人数量足够大的情况,在以往有关多属性拍卖研究的基础上将单个投入扩展为多个投入,将物品的多个属性(价格、完成时间、劳动力数量、质量等因素)合理划分为投入或产出,设计了运行有效的基于数据包络分析的多属性第二分值拍卖机制。与其他方法相比,数据包络分析方法可以有效解决多投入多产出问题,结合多目标规划方法,可以帮助采购方在最大化自身利益的同时,找到整体表现更好的供应商。该机制满足个人理性和激励相容,与传统第二分值拍卖机制相比,在吸引投标人的前提下能够最大化拍卖人的利益。

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, the new mechanism can maximize the utility of the auctioneer under the premise of attracting bidders.