Learning to Rank with Top-$K$ Fairness

2025-09-22

AI Summary

Existing fair ranking methods primarily optimize average exposure across the entire ranked list, overlooking the practical requirement that decision-makers often focus exclusively on the top-K items. This paper introduces the first Top-K–aware fair ranking framework, which directly optimizes the trade-off between group fairness and relevance within the top-K positions during training. Its core contributions are: (1) a differentiable Top-K exposure disparity metric that reformulates the non-differentiable Top-K selection as a continuous optimization objective; and (2) an efficient learning-to-rank approach at the list level, integrating differentiable approximations with stochastic optimization. Experiments demonstrate that the proposed method achieves high ranking accuracy while significantly reducing unfair exposure in the top-K positions, consistently outperforming state-of-the-art fair ranking baselines across multiple fairness and utility metrics.

Technology Category

• Machine Learning: Learning Preferences or Rankings
• Search and Optimization: Learning to Search
• Computer Vision: Bias, Fairness & Privacy

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Algorithmic accountability and transparency on the web

Abstract

Fairness in ranking models is crucial, as disparities in exposure can disproportionately affect protected groups. Most fairness-aware ranking systems focus on ensuring comparable average exposure for groups across the entire ranked list, which may not fully address real-world concerns. For example, when a ranking model is used for allocating resources among candidates or disaster hotspots, decision-makers often prioritize only the top-$K$ ranked items, while the ranking beyond top-$K$ becomes less relevant. In this paper, we propose a list-wise learning-to-rank framework that addresses the issues of inequalities in top-$K$ rankings at training time. Specifically, we propose a top-$K$ exposure disparity measure that extends the classic exposure disparity metric in a ranked list. We then learn a ranker to balance relevance and fairness in top-$K$ rankings. Since direct top-$K$ selection is computationally expensive for a large number of items, we transform the non-differentiable selection process into a differentiable objective function and develop efficient stochastic optimization algorithms to achieve both high accuracy and sufficient fairness. Extensive experiments demonstrate that our method outperforms existing methods.

Problem

• Addresses fairness disparities in top-K ranked item exposure
• Balances relevance and fairness specifically for top-K rankings
• Transforms non-differentiable top-K selection into differentiable optimization

Innovation

• Top-K fairness ranking framework
• Differentiable top-K selection objective
• Stochastic optimization for accuracy and fairness