Personalized One-shot Federated Graph Learning for Heterogeneous Clients

TL;DR

This paper introduces O-pFGL, a novel one-shot federated graph learning method that efficiently personalizes models for heterogeneous clients, reducing communication rounds and enhancing privacy in node classification tasks.

Contribution

It presents the first one-shot personalized federated graph learning approach that constructs a global surrogate graph and balances local and global information for improved personalization.

Findings

Outperforms state-of-the-art methods on 14 real-world datasets.

Effectively reduces communication overhead in federated graph learning.

Enhances model generalization and personalization for heterogeneous clients.

Abstract

Federated Graph Learning (FGL) has emerged as a promising paradigm for breaking data silos among distributed private graphs. In practical scenarios involving heterogeneous distributed graph data, personalized Federated Graph Learning (pFGL) aims to enhance model utility by training personalized models tailored to client needs. However, existing pFGL methods often require numerous communication rounds under heterogeneous graphs, leading to significant communication overhead and security concerns. While One-shot Federated Learning (OFL) enables collaboration in a single round, existing OFL methods are designed for image-centric tasks and are ineffective for graph data, leaving a critical gap in the field. Additionally, personalized models derived from existing methods suffer from bias, failing to effectively generalize to the minority. To address these challenges, we propose the first \textbf{O}ne-shot \textbf{p}ersonalized \textbf{F}ederated \textbf{G}raph \textbf{L}earning method (\textbf{O-pFGL}) for node classification, compatible with Secure Aggregation protocols for privacy preservation. Specifically, for effective graph learning in one communication round, our method estimates and aggregates class-wise feature distribution statistics to construct a global surrogate graph on the server, facilitating the training of a global graph model. To mitigate bias, we introduce a two-stage personalized training approach that adaptively balances local personal information and global insights from the surrogate graph, improving both personalization and generalization. Extensive experiments on 14 diverse real-world graph datasets demonstrate that our method significantly outperforms state-of-the-art baselines across various settings.