Toward Unifying Group Fairness Evaluation from a Sparsity Perspective

2025-10-31

AI Summary

Existing fairness evaluation metrics suffer from limited generalizability across tasks and weak interpretability. Method: This paper proposes a unified fairness evaluation framework grounded in sparsity theory, establishing the first formal theoretical connection between social fairness and the sparsity of model parameters or prediction distributions. It reformulates over a dozen mainstream group fairness criteria—including statistical parity and equal opportunity—as optimization problems subject to sparsity constraints. A differentiable sparsity metric is designed and integrated with sensitive attribute analysis to enable end-to-end, quantitative fairness assessment across diverse datasets and bias-mitigation algorithms. Results: Experiments on eight benchmark datasets and six classes of debiasing methods demonstrate that the framework achieves strong generalizability, stability, and interpretability, significantly improving consistency and practical utility in fairness evaluation.

Technology Category

• Computer Vision: Bias, Fairness & Privacy
• Machine Learning: Ethics, Bias, and Fairness
• Philosophy and Ethics of AI: Bias, Fairness & Equity

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Social Networks and Social Media: Fairness and bias in social network and social media analysis
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Ensuring algorithmic fairness remains a significant challenge in machine learning, particularly as models are increasingly applied across diverse domains. While numerous fairness criteria exist, they often lack generalizability across different machine learning problems. This paper examines the connections and differences among various sparsity measures in promoting fairness and proposes a unified sparsity-based framework for evaluating algorithmic fairness. The framework aligns with existing fairness criteria and demonstrates broad applicability to a wide range of machine learning tasks. We demonstrate the effectiveness of the proposed framework as an evaluation metric through extensive experiments on a variety of datasets and bias mitigation methods. This work provides a novel perspective to algorithmic fairness by framing it through the lens of sparsity and social equity, offering potential for broader impact on fairness research and applications.

Problem

• Unifying sparsity measures for algorithmic fairness evaluation
• Proposing a sparsity-based framework for fairness across ML tasks
• Addressing fairness generalizability through sparsity and social equity

Innovation

• Unified sparsity framework for fairness evaluation
• Aligns with existing fairness criteria broadly
• Frames fairness through sparsity and equity lens