Let's Grow an Unbiased Community: Guiding the Fairness of Graphs via New Links

TL;DR

This paper introduces FairGuide, a framework that guides the structure of graphs towards fairness by adding new links, using a differentiable community detection task and meta-gradient strategies to improve fairness in graph neural network applications.

Contribution

The paper proposes a novel fairness-guided link addition framework for graphs, employing a differentiable community detection pseudo task and meta-gradient optimization to enhance structural fairness.

Findings

FairGuide effectively improves fairness in various graph-based tasks.

The method demonstrates strong generalizability across different datasets.

Experimental results show significant fairness enhancement without sacrificing accuracy.

Abstract

Graph Neural Networks (GNNs) have achieved remarkable success across diverse applications. However, due to the biases in the graph structures, graph neural networks face significant challenges in fairness. Although the original user graph structure is generally biased, it is promising to guide these existing structures toward unbiased ones by introducing new links. The fairness guidance via new links could foster unbiased communities, thereby enhancing fairness in downstream applications. To address this issue, we propose a novel framework named FairGuide. Specifically, to ensure fairness in downstream tasks trained on fairness-guided graphs, we introduce a differentiable community detection task as a pseudo downstream task. Our theoretical analysis further demonstrates that optimizing fairness within this pseudo task effectively enhances structural fairness, promoting fairness generalization across diverse downstream applications. Moreover, FairGuide employs an effective strategy which leverages meta-gradients derived from the fairness-guidance objective to identify new links that significantly enhance structural fairness. Extensive experimental results demonstrate the effectiveness and generalizability of our proposed method across a variety of graph-based fairness tasks.