Fairness-Aware Graph Representation Learning with Limited Demographic Information
Zichong Wang, Zhipeng Yin, Liping Yang, Jun Zhuang, Rui Yu, Qingzhao Kong, Wenbin Zhang
TL;DR
This work tackles fairness in graph neural networks when demographic information is incomplete by introducing FairGLite, a proxy-based, causally motivated framework. It learns demographic proxies via a dedicated encoder, enforces fairness through a trio of constraints (masking-based parity, information preservation, and ego-graph reconstruction), and applies an adaptive confidence strategy to weight fairness more heavily on high-confidence samples while maintaining utility. The method is theoretically grounded, providing upper bounds on group-fairness metrics, and empirically validated on four real-world datasets where 40% of demographic labels are masked, showing improved bias mitigation with competitive predictive performance. Overall, FairGLite advances practical fair graph learning in privacy- and regulation-constrained settings and offers formal guarantees for bias reduction in downstream tasks.
Abstract
Ensuring fairness in Graph Neural Networks is fundamental to promoting trustworthy and socially responsible machine learning systems. In response, numerous fair graph learning methods have been proposed in recent years. However, most of them assume full access to demographic information, a requirement rarely met in practice due to privacy, legal, or regulatory restrictions. To this end, this paper introduces a novel fair graph learning framework that mitigates bias in graph learning under limited demographic information. Specifically, we propose a mechanism guided by partial demographic data to generate proxies for demographic information and design a strategy that enforces consistent node embeddings across demographic groups. In addition, we develop an adaptive confidence strategy that dynamically adjusts each node's contribution to fairness and utility based on prediction confidence. We further provide theoretical analysis demonstrating that our framework, FairGLite, achieves provable upper bounds on group fairness metrics, offering formal guarantees for bias mitigation. Through extensive experiments on multiple datasets and fair graph learning frameworks, we demonstrate the framework's effectiveness in both mitigating bias and maintaining model utility.