On Ordered Weighted Reference Point Model for Multi-attribute Procurement Auctions

Multi-attribute auctions (also called multidimensional auctions) facilitate negotiations based on multiple attributes, thus escape from the standard price-only domain into a rich multidimensional domain that can comprise additional attributes like e.g. guarantee conditions or quality. Most multi-attribute preference models used in auction mechanisms are based on a ranking derived from weighted sum. Recently, the Reference Point Method (RPM) approach has been applied to express the multi-attribute preference models within the auction mechanisms allowing to overcome the weighted sum drawbacks. The Ordered Weighted RPM model enables us to introduce importance weights to affect achievements importance by rescaling their measures accordingly within the distribution of all achievements. The concept presented and discussed in this paper in the context of procurement auctions takes advantage of the so-called Weighted Ordered Weighted Average (WOWA) aggregations of the partial achievements.