Homophily-aware Supervised Contrastive Counterfactual Augmented Fair Graph Neural Network

TL;DR

This paper introduces a fairness-aware GNN model that enhances predictive accuracy and fairness by editing graph homophily and employing a novel contrastive loss in a two-phase training process.

Contribution

It proposes a new two-phase training strategy that edits graph homophily and integrates a modified contrastive loss to improve fairness and accuracy in GNNs.

Findings

Outperforms existing methods in accuracy and fairness on five datasets.

Effectively increases homophily for class labels while decreasing it for sensitive attributes.

Demonstrates the effectiveness of the two-phase training strategy.

Abstract

In recent years, Graph Neural Networks (GNNs) have achieved remarkable success in tasks such as node classification, link prediction, and graph representation learning. However, they remain susceptible to biases that can arise not only from node attributes but also from the graph structure itself. Addressing fairness in GNNs has therefore emerged as a critical research challenge. In this work, we propose a novel model for training fairness-aware GNNs by improving the counterfactual augmented fair graph neural network framework (CAF). Specifically, our approach introduces a two-phase training strategy: in the first phase, we edit the graph to increase homophily ratio with respect to class labels while reducing homophily ratio with respect to sensitive attribute labels; in the second phase, we integrate a modified supervised contrastive loss and environmental loss into the optimization process, enabling the model to jointly improve predictive performance and fairness. Experiments on five real-world datasets demonstrate that our model outperforms CAF and several state-of-the-art graph-based learning methods in both classification accuracy and fairness metrics.