Policy Learning for Fairness in Ranking

TL;DR

This paper introduces a flexible learning-to-rank framework that optimizes utility while ensuring fairness in exposure, using stochastic policies and a new policy-gradient algorithm, with demonstrated effectiveness on real and simulated data.

Contribution

It presents a novel LTR framework that incorporates fairness constraints and a new algorithm, Fair-PG-Rank, for learning fair ranking policies.

Findings

Effective in achieving fairness in exposure for items.

Works on both simulated and real-world datasets.

Balances utility and fairness successfully.

Abstract

Conventional Learning-to-Rank (LTR) methods optimize the utility of the rankings to the users, but they are oblivious to their impact on the ranked items. However, there has been a growing understanding that the latter is important to consider for a wide range of ranking applications (e.g. online marketplaces, job placement, admissions). To address this need, we propose a general LTR framework that can optimize a wide range of utility metrics (e.g. NDCG) while satisfying fairness of exposure constraints with respect to the items. This framework expands the class of learnable ranking functions to stochastic ranking policies, which provides a language for rigorously expressing fairness specifications. Furthermore, we provide a new LTR algorithm called Fair-PG-Rank for directly searching the space of fair ranking policies via a policy-gradient approach. Beyond the theoretical evidence in deriving the framework and the algorithm, we provide empirical results on simulated and real-world datasets verifying the effectiveness of the approach in individual and group-fairness settings.