Mitigating topology biases in Graph Diffusion via Counterfactual Intervention

TL;DR

This paper introduces FairGDiff, a counterfactual-based graph diffusion model that reduces topology biases related to sensitive attributes, improving fairness without sacrificing utility in graph generation.

Contribution

The paper presents a novel counterfactual intervention method applied directly to graph topology within a diffusion model, addressing limitations of prior fair graph generation approaches.

Findings

Outperforms existing methods in fairness-utility trade-off.

Effectively mitigates topology biases related to sensitive attributes.

Maintains scalability and structural integrity in generated graphs.

Abstract

Graph diffusion models have gained significant attention in graph generation tasks, but they often inherit and amplify topology biases from sensitive attributes (e.g. gender, age, region), leading to unfair synthetic graphs. Existing fair graph generation using diffusion models is limited to specific graph-based applications with complete labels or requires simultaneous updates for graph structure and node attributes, making them unsuitable for general usage. To relax these limitations by applying the debiasing method directly on graph topology, we propose Fair Graph Diffusion Model (FairGDiff), a counterfactual-based one-step solution that mitigates topology biases while balancing fairness and utility. In detail, we construct a causal model to capture the relationship between sensitive attributes, biased link formation, and the generated graph structure. By answering the counterfactual question "Would the graph structure change if the sensitive attribute were different?", we estimate an unbiased treatment and incorporate it into the diffusion process. FairGDiff integrates counterfactual learning into both forward diffusion and backward denoising, ensuring that the generated graphs are independent of sensitive attributes while preserving structural integrity. Extensive experiments on real-world datasets demonstrate that FairGDiff achieves a superior trade-off between fairness and utility, outperforming existing fair graph generation methods while maintaining scalability.