Fairness in Combinatorial Auctioning Systems

TL;DR

This paper explores fairness in combinatorial auctioning systems, proposing algorithms and mathematical models to ensure fairness while maintaining optimal resource allocation using VCG mechanisms.

Contribution

It introduces two fairness notions in CAS, develops algorithms incorporating fairness metrics, and provides mathematical formulations for fairness measures.

Findings

Algorithms successfully incorporate fairness metrics into CAS.

Mathematical models quantify extended and basic fairness.

The approach maintains optimality with VCG mechanisms.

Abstract

One of the Multi-Agent Systems that is widely used by various government agencies, buyers and sellers in a market economy, in such a manner so as to attain optimized resource allocation, is the Combinatorial Auctioning System (CAS). We study another important aspect of resource allocations in CAS, namely fairness. We present two important notions of fairness in CAS, extended fairness and basic fairness. We give an algorithm that works by incorporating a metric to ensure fairness in a CAS that uses the Vickrey-Clark-Groves (VCG) mechanism, and uses an algorithm of Sandholm to achieve optimality. Mathematical formulations are given to represent measures of extended fairness and basic fairness.