FairContrast: Enhancing Fairness through Contrastive learning and Customized Augmenting Methods on Tabular Data

2025-10-02

AI Summary

Existing self-supervised and contrastive learning methods for tabular data representation learning inadequately address fairness, lacking explicit debiasing mechanisms. To bridge this gap, we propose F-TabCLR, a fairness-aware contrastive learning framework for tabular data. F-TabCLR introduces a sensitive-attribute-aware positive sample selection strategy and integrates supervised signals with tailored tabular augmentations—including numerical perturbation and class reweighting—to explicitly disentangle sensitive information from task-relevant features in the representation space. Extensive experiments on multiple real-world tabular datasets demonstrate that F-TabCLR significantly reduces group- and individual-level fairness disparities (e.g., ΔSP and ΔEO improve by 15–32%), while maintaining or even enhancing downstream predictive accuracy. To our knowledge, F-TabCLR is the first framework to achieve joint optimization of fairness and prediction performance in tabular representation learning.

Technology Category

• Machine Learning: Representation Learning
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy
• Computer Vision: Representation Learning for Vision

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
• Economics, Online Markets and Human Computation: Fairness, privacy, and diversity in economic environments

Abstract

As AI systems become more embedded in everyday life, the development of fair and unbiased models becomes more critical. Considering the social impact of AI systems is not merely a technical challenge but a moral imperative. As evidenced in numerous research studies, learning fair and robust representations has proven to be a powerful approach to effectively debiasing algorithms and improving fairness while maintaining essential information for prediction tasks. Representation learning frameworks, particularly those that utilize self-supervised and contrastive learning, have demonstrated superior robustness and generalizability across various domains. Despite the growing interest in applying these approaches to tabular data, the issue of fairness in these learned representations remains underexplored. In this study, we introduce a contrastive learning framework specifically designed to address bias and learn fair representations in tabular datasets. By strategically selecting positive pair samples and employing supervised and self-supervised contrastive learning, we significantly reduce bias compared to existing state-of-the-art contrastive learning models for tabular data. Our results demonstrate the efficacy of our approach in mitigating bias with minimum trade-off in accuracy and leveraging the learned fair representations in various downstream tasks.

Problem

• Addressing bias in tabular data using contrastive learning methods
• Developing fair representations while maintaining prediction task accuracy
• Reducing algorithmic bias through customized sample selection strategies

Innovation

• Uses contrastive learning for tabular data fairness
• Selects positive pairs strategically to reduce bias
• Combines supervised and self-supervised contrastive learning methods