On Graph Neural Network Fairness in the Presence of Heterophilous Neighborhoods

TL;DR

This paper explores how local neighborhood structures in graphs affect fairness in GNN-based node classification, revealing that heterophilous designs can improve fairness in diverse neighborhoods.

Contribution

It establishes a link between local homophily and fairness, and demonstrates that heterophilous GNNs enhance fairness in heterophilous neighborhoods.

Findings

Heterophilous GNNs improve fairness by up to 25% in diverse neighborhoods.

Neighborhood homophily influences the fairness of GNN predictions.

Designing GNNs for heterophily can mitigate unfair treatment in sensitive attribute classification.

Abstract

We study the task of node classification for graph neural networks (GNNs) and establish a connection between group fairness, as measured by statistical parity and equal opportunity, and local assortativity, i.e., the tendency of linked nodes to have similar attributes. Such assortativity is often induced by homophily, the tendency for nodes of similar properties to connect. Homophily can be common in social networks where systemic factors have forced individuals into communities which share a sensitive attribute. Through synthetic graphs, we study the interplay between locally occurring homophily and fair predictions, finding that not all node neighborhoods are equal in this respect -- neighborhoods dominated by one category of a sensitive attribute often struggle to obtain fair treatment, especially in the case of diverging local class and sensitive attribute homophily. After determining that a relationship between local homophily and fairness exists, we investigate if the issue of unfairness can be associated to the design of the applied GNN model. We show that by adopting heterophilous GNN designs capable of handling disassortative group labels, group fairness in locally heterophilous neighborhoods can be improved by up to 25% over homophilous designs in real and synthetic datasets.