Subgraph Federated Learning for Local Generalization

2025-03-06

AI Summary

In graph federated learning, local overfitting severely degrades model generalization to unseen nodes and non-stationary label distributions. To address this, we propose FedLoG—the first framework to introduce a global synthetic data mechanism into graph federated learning. FedLoG aggregates class-level representations and structural information across clients to generate globally synthetic graph data that is both discriminative and topologically consistent. It further incorporates a structure-aware class representation compression strategy to enable efficient knowledge distillation and enhanced generalization. Evaluated on multiple dynamic graph benchmarks, FedLoG significantly improves adaptability to unseen nodes and distribution shifts, achieving average accuracy gains of 5.2%–9.7% over state-of-the-art methods. The implementation is publicly available.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Search and Optimization: Local Search

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Federated Learning (FL) on graphs enables collaborative model training to enhance performance without compromising the privacy of each client. However, existing methods often overlook the mutable nature of graph data, which frequently introduces new nodes and leads to shifts in label distribution. Since they focus solely on performing well on each client's local data, they are prone to overfitting to their local distributions (i.e., local overfitting), which hinders their ability to generalize to unseen data with diverse label distributions. In contrast, our proposed method, FedLoG, effectively tackles this issue by mitigating local overfitting. Our model generates global synthetic data by condensing the reliable information from each class representation and its structural information across clients. Using these synthetic data as a training set, we alleviate the local overfitting problem by adaptively generalizing the absent knowledge within each local dataset. This enhances the generalization capabilities of local models, enabling them to handle unseen data effectively. Our model outperforms baselines in our proposed experimental settings, which are designed to measure generalization power to unseen data in practical scenarios. Our code is available at https://github.com/sung-won-kim/FedLoG

Problem

• Addresses local overfitting in federated learning on graphs
• Enhances generalization to unseen data with diverse label distributions
• Mitigates issues from mutable graph data and new node introductions

Innovation

• Generates global synthetic data from class representations
• Uses synthetic data to mitigate local overfitting
• Enhances generalization to unseen data effectively