Learning Fair Graph Representations with Multi-view Information Bottleneck

TL;DR

FairMIB is a novel multi-view information bottleneck framework that enhances fairness in graph neural networks by decomposing graph information, employing contrastive learning, and correcting bias propagation, achieving state-of-the-art results.

Contribution

It introduces a multi-view approach with contrastive learning and IPW adjacency correction to improve fairness and utility in GNNs, addressing bias propagation more effectively.

Findings

Achieves state-of-the-art fairness and utility on benchmark datasets.

Effectively mitigates bias propagation in graph message passing.

Balances fairness and utility through multi-perspective objectives.

Abstract

Graph neural networks (GNNs) excel on relational data by passing messages over node features and structure, but they can amplify training data biases, propagating discriminatory attributes and structural imbalances into unfair outcomes. Many fairness methods treat bias as a single source, ignoring distinct attribute and structure effects and leading to suboptimal fairness and utility trade-offs. To overcome this challenge, we propose FairMIB, a multi-view information bottleneck framework designed to decompose graphs into feature, structural, and diffusion views for mitigating complexity biases in GNNs. Especially, the proposed FairMIB employs contrastive learning to maximize cross-view mutual information for bias-free representation learning. It further integrates multi-perspective conditional information bottleneck objectives to balance task utility and fairness by minimizing mutual information with sensitive attributes. Additionally, FairMIB introduces an inverse probability-weighted (IPW) adjacency correction in the diffusion view, which reduces the spread of bias propagation during message passing. Experiments on five real-world benchmark datasets demonstrate that FairMIB achieves state-of-the-art performance across both utility and fairness metrics.