Fair Mechanisms for Recurrent Multi Unit Combinatorial Auctions

  • Authors:

  • Javier Murillo;Beatriz López

  • Affiliations:

  • University of Girona, Spain;University of Girona, Spain

  • Venue:
  • Proceedings of the 2010 conference on STAIRS 2010: Proceedings of the Fifth Starting AI Researchers' Symposium
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

Auctions have been used to deal with resource allocation in multi-agent systems. In some environments like service-oriented electronic markets, it is advisable to use recurrent auctions since resources are perishable and auctions are repeated over time with the same or a very similar set of agents. Recurrent auctions are a series of auctions of any kind where the result of one auction may influence the following one. As a drawback some problems do appear that could cause the market to collapse at mid-long term. Previous works have dealt with these problems by adding fairness to the auction outcomes. Those works dealt with multi-unit auctions, in which several units of an item are sold, and they do not assure that agents cannot manipulate the auctions for their own benefit. In this paper, we present new fair mechanisms that goes further. First we focus on combinatorial auctions, in which different items, and several units per item are sold in each auction, which poses additional challenges when they are recurrent. And second, the mechanisms are shown to prevent some agents' manipulation of the auction outcomes.