A fairness-aware extension of Stochastic Multicriteria Acceptability Analysis for ranking

2026-06-16

AI Summary

This study addresses the frequent neglect of group fairness in traditional multicriteria ranking methods under preference uncertainty, which often leads to underrepresentation of disadvantaged groups. To bridge this gap, the authors propose SMAA-Fair, the first approach to embed fairness mechanisms directly into the Stochastic Multicriteria Acceptability Analysis (SMAA) framework. SMAA-Fair reweights simulated rankings according to fairness metrics—namely statistical parity, normalized discounted KL divergence (rKL), and nDKL—so that fairer rankings receive higher weight in both acceptability indices and central weights. Notably, the method is agnostic to any specific aggregation model, thereby preserving robustness while promoting fairness. Experimental results demonstrate that SMAA-Fair significantly enhances the representation of protected groups in top ranking positions without compromising robustness to preference uncertainty.

Abstract

Fairness has become a central concern in ranking problems involving individuals or social groups, particularly under the Responsible Artificial Intelligence agenda. In Multi-Criteria Decision Analysis, Stochastic Multicriteria Acceptability Analysis (SMAA) provides a robust framework for handling uncertainty and incomplete preference information, but it does not explicitly address fairness in the resulting rankings. This paper proposes SMAA-Fair, a fairness-aware extension of SMAA for ranking problems. The approach reweights the simulated rankings generated by SMAA according to their level of group fairness, so that fairer rankings contribute more strongly to the acceptability indices and central weights vector. The framework is independent of the aggregation model and can incorporate different fairness metrics. In this study, Statistical Parity, normalized discounted Kullback--Leibler divergence (rKL) and normalized discounted cumulative Kullback--Leibler divergence (nDKL) are adopted. Rankings are derived from the fairness-adjusted acceptability matrix using expected ranking and maximum acceptability ranking. We also derive the central weight according to the degree of fairness in the obtained rankings. Numerical experiments with synthetic and real data show that SMAA-Fair improves the representation of protected groups among favourable ranking positions, while preserving robustness to preference uncertainty.

Problem

• fairness
• ranking
• Stochastic Multicriteria Acceptability Analysis
• group fairness
• preference uncertainty

Innovation

• fairness-aware ranking
• Stochastic Multicriteria Acceptability Analysis
• group fairness
• preference uncertainty
• SMAA-Fair