FedSDR: Federated Graph Learning with Structural Noise Detection and Reconstruction

Jiaqi Liu, Zihan Tan, Guancheng Wan, Wenke Huang, He Li, Mang Ye

Keywords: Federated Learning; Graph Learning

Abstract: Federated Graph Learning (FGL) has emerged as a principled framework for decentralized training of Graph Neural Networks (GNNs) while preserving data privacy. In subgraph-FL scenarios, however, structural noise arising from data collection and storage can damage the GNN message-passing scheme of clients, leading to conflicts in collaboration. Existing approaches exhibit two critical limitations: 1) Globally, they fail to identify corrupted clients, causing destructive message-passing conflicts. 2) Locally, the global GNN performs poorly on these clients due to structural noise, limiting their ability to benefit from federated collaboration. To address these challenges, we propose $\textbf{FedSDR}$, a robust FGL framework against high-structural-noise scenarios. Specifically, Structural Noise-Aware Aggregation (SNAA) introduces a noise evaluation metric to detect corrupted clients and reduce their contributions, thereby mitigating the impact of noise on the global GNN. Furthermore, Robust Local Structure Reconstruction (RLSR) leverages the knowledge from the healthy global model to repair locally corrupted graph structures. Extensive experiments demonstrate that FedSDR outperforms state-of-the-art methods across various scenarios under structural noise.

Primary Area: alignment, fairness, safety, privacy, and societal considerations

Submission Number: 3767