GraphFedMIG: Tackling Class Imbalance in Federated Graph Learning via Mutual Information-Guided Generation

2025-08-14

AI Summary

In federated graph learning (FGL), non-IID client data induces class imbalance, degrading global model performance on minority classes and biasing neighborhood aggregation—thereby impairing embedding expressiveness. To address this, we propose a mutual information-guided generative data augmentation framework: a hierarchical GAN architecture integrating client-cluster discriminators and local generators; and a novel mutual information maximization mechanism that dynamically quantifies client-specific information value to steer the generator toward synthesizing high-value minority-class features. Extensive experiments on four real-world graph datasets demonstrate that our method significantly outperforms state-of-the-art baselines in both classification accuracy and fairness metrics—including minority-class F1-score and group fairness—effectively mitigating model bias under non-IID conditions.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Social Networks and Social Media: Fairness and bias in social network and social media analysis

Abstract

Federated graph learning (FGL) enables multiple clients to collaboratively train powerful graph neural networks without sharing their private, decentralized graph data. Inherited from generic federated learning, FGL is critically challenged by statistical heterogeneity, where non-IID data distributions across clients can severely impair model performance. A particularly destructive form of this is class imbalance, which causes the global model to become biased towards majority classes and fail at identifying rare but critical events. This issue is exacerbated in FGL, as nodes from a minority class are often surrounded by biased neighborhood information, hindering the learning of expressive embeddings. To grapple with this challenge, we propose GraphFedMIG, a novel FGL framework that reframes the problem as a federated generative data augmentation task. GraphFedMIG employs a hierarchical generative adversarial network where each client trains a local generator to synthesize high-fidelity feature representations. To provide tailored supervision, clients are grouped into clusters, each sharing a dedicated discriminator. Crucially, the framework designs a mutual information-guided mechanism to steer the evolution of these client generators. By calculating each client's unique informational value, this mechanism corrects the local generator parameters, ensuring that subsequent rounds of mutual information-guided generation are focused on producing high-value, minority-class features. We conduct extensive experiments on four real-world datasets, and the results demonstrate the superiority of the proposed GraphFedMIG compared with other baselines.

Problem

• Addresses class imbalance in federated graph learning
• Mitigates bias from non-IID data across clients
• Generates minority-class features via mutual information guidance

Innovation

• Hierarchical GAN for federated feature generation
• Mutual information-guided client parameter correction
• Cluster-shared discriminators for tailored supervision