One Fits All: Learning Fair Graph Neural Networks for Various Sensitive Attributes

TL;DR

This paper introduces FairINV, a novel invariant learning framework for GNNs that achieves fairness across multiple sensitive attributes in a single training process, reducing computational costs and improving fairness performance.

Contribution

The paper proposes a new invariant learning-based framework, FairINV, enabling fair GNN training for various sensitive attributes without retraining from scratch.

Findings

FairINV outperforms existing fairness methods on real-world datasets.

It effectively eliminates spurious correlations between labels and sensitive attributes.

The approach reduces the need for multiple retraining sessions for different attributes.

Abstract

Recent studies have highlighted fairness issues in Graph Neural Networks (GNNs), where they produce discriminatory predictions against specific protected groups categorized by sensitive attributes such as race and age. While various efforts to enhance GNN fairness have made significant progress, these approaches are often tailored to specific sensitive attributes. Consequently, they necessitate retraining the model from scratch to accommodate changes in the sensitive attribute requirement, resulting in high computational costs. To gain deeper insights into this issue, we approach the graph fairness problem from a causal modeling perspective, where we identify the confounding effect induced by the sensitive attribute as the underlying reason. Motivated by this observation, we formulate the fairness problem in graphs from an invariant learning perspective, which aims to learn invariant representations across environments. Accordingly, we propose a graph fairness framework based on invariant learning, namely FairINV, which enables the training of fair GNNs to accommodate various sensitive attributes within a single training session. Specifically, FairINV incorporates sensitive attribute partition and trains fair GNNs by eliminating spurious correlations between the label and various sensitive attributes. Experimental results on several real-world datasets demonstrate that FairINV significantly outperforms state-of-the-art fairness approaches, underscoring its effectiveness. Our code is available via: https://github.com/ZzoomD/FairINV/.