Post-hoc Provider Fairness Adaptation via Hierarchical Exposure Alignment

2026-05-02

AI Summary

Existing recommender systems often struggle to achieve provider-side exposure fairness due to high retraining costs or inflexibility in accommodating diverse fairness requirements. This work proposes PFA, a lightweight post-processing framework that enables flexible fairness control without model retraining by freezing the main recommendation model and introducing a fairness adapter. The core innovation lies in the Hierarchical Exposure Fairness Alignment (HEFA) mechanism, which explicitly co-optimizes both inter-group and intra-group exposure equity among providers. PFA further incorporates personalized additive adjustments based on user-item embeddings, a KL-divergence-based optimization objective, and a differentiable NDCG loss. Experiments on three public datasets demonstrate that PFA significantly improves multiple fairness metrics while preserving recommendation accuracy almost perfectly, outperforming strong existing baselines across the board.

Abstract

Provider exposure fairness is crucial for sustaining a healthy content ecosystem and preventing monopolization in recommender systems. Yet, most existing methods either incorporate fairness constraints during model training, requiring expensive retraining when fairness objectives change, or rely on post-hoc reranking with fixed criteria, which lacks adaptability to diverse fairness requirements. To overcome these limitations, we propose Post-hoc Fairness Adaptation (PFA), a lightweight framework that equips a frozen recommender with a fairness adapter, enabling flexible fairness control without retraining the backbone model. Specifically, the fairness adapter learns personalized additive score adjustments from user-item embeddings, which are injected into the original ranking scores to steer provider exposure toward fairness. To train the adapter, we minimize the KL divergence between the actual and the target fair exposure distributions. However, this global objective implicitly treats all providers equally, ignoring structural disparities such as imbalanced provider group sizes and heterogeneous exposure within groups. Consequently, fairness may appear satisfied at an aggregate level while severe inter-group and intra-group exposure imbalances persist, undermining practical fairness. To address this, we design Hierarchical Exposure Fairness Alignment (HEFA), which explicitly balances inter- and intra-group provider exposure disparities, enabling flexible adaptation to diverse fairness requirements. To mitigate potential accuracy degradation, PFA jointly optimizes HEFA with a differentiable NDCG loss, enabling end-to-end fairness optimization while preserving ranking quality. Extensive experiments on three public datasets demonstrate that PFA achieves substantial fairness gains with negligible accuracy loss, consistently outperforming strong baselines.

Problem

• provider fairness
• exposure imbalance
• post-hoc adaptation
• hierarchical fairness
• recommender systems

Innovation

• Post-hoc Fairness Adaptation
• Hierarchical Exposure Alignment
• Provider Fairness
• Recommender Systems
• Exposure Disparity