FedHERO: A Federated Learning Approach for Node Classification Task on Heterophilic Graphs

2025-04-29

AI Summary

To address the severe performance degradation in federated graph learning (FGL) caused by highly heterogeneous neighborhood distributions across client graphs—which induce local model knowledge conflicts and hinder effective global aggregation—this paper proposes the first dual-channel GNN framework tailored for heterogeneous graphs. The framework jointly couples a structure learner and a feature learner: the former optimizes graph topology via learnable structural refinement to extract cross-graph generalizable topological knowledge, while the latter incorporates a heterogeneity-aware federated aggregation mechanism that ensures robust collaborative training under node-level differential privacy guarantees. Evaluated on multiple heterogeneous graph benchmarks, the approach improves global model accuracy by an average of 12.6%, simultaneously enhancing both local model generalization and robustness. This work establishes the first effective solution for federated graph representation learning over multi-modal neighborhood-distribution graphs.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Search and Optimization: Learning to Search
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Federated Graph Learning (FGL) empowers clients to collaboratively train Graph neural networks (GNNs) in a distributed manner while preserving data privacy. However, FGL methods usually require that the graph data owned by all clients is homophilic to ensure similar neighbor distribution patterns of nodes. Such an assumption ensures that the learned knowledge is consistent across the local models from all clients. Therefore, these local models can be properly aggregated as a global model without undermining the overall performance. Nevertheless, when the neighbor distribution patterns of nodes vary across different clients (e.g., when clients hold graphs with different levels of heterophily), their local models may gain different and even conflict knowledge from their node-level predictive tasks. Consequently, aggregating these local models usually leads to catastrophic performance deterioration on the global model. To address this challenge, we propose FedHERO, an FGL framework designed to harness and share insights from heterophilic graphs effectively. At the heart of FedHERO is a dual-channel GNN equipped with a structure learner, engineered to discern the structural knowledge encoded in the local graphs. With this specialized component, FedHERO enables the local model for each client to identify and learn patterns that are universally applicable across graphs with different patterns of node neighbor distributions. FedHERO not only enhances the performance of individual client models by leveraging both local and shared structural insights but also sets a new precedent in this field to effectively handle graph data with various node neighbor distribution patterns. We conduct extensive experiments to validate the superior performance of FedHERO against existing alternatives.

Problem

• Addresses performance drop in federated learning with heterophilic graphs
• Enables effective knowledge sharing across diverse node distribution patterns
• Proposes dual-channel GNN to handle varying neighbor distribution structures

Innovation

• Federated learning for heterophilic graph classification
• Dual-channel GNN with structure learner
• Shares universally applicable structural patterns