HoGS: Homophily-Oriented Graph Synthesis for Local Differentially Private GNN Training

TL;DR

HoGS introduces a novel framework that synthesizes graphs under local differential privacy by leveraging homophily, enabling effective GNN training with enhanced privacy and accuracy on real-world datasets.

Contribution

HoGS presents a new LDP-based graph synthesis method that preserves link and feature privacy while maintaining high GNN performance, addressing limitations of existing approaches.

Findings

HoGS achieves higher GNN accuracy than baseline methods.

The framework effectively balances privacy and utility in graph data.

Experimental results validate the theoretical privacy guarantees.

Abstract

Graph neural networks (GNNs) have demonstrated remarkable performance in various graph-based machine learning tasks by effectively modeling high-order interactions between nodes. However, training GNNs without protection may leak sensitive personal information in graph data, including links and node features. Local differential privacy (LDP) is an advanced technique for protecting data privacy in decentralized networks. Unfortunately, existing local differentially private GNNs either only preserve link privacy or suffer significant utility loss in the process of preserving link and node feature privacy. In this paper, we propose an effective LDP framework, called HoGS, which trains GNNs with link and feature protection by generating a synthetic graph. Concretely, HoGS first collects the link and feature information of the graph under LDP, and then utilizes the phenomenon of homophily in graph data to reconstruct the graph structure and node features separately, thereby effectively mitigating the negative impact of LDP on the downstream GNN training. We theoretically analyze the privacy guarantee of HoGS and conduct experiments using the generated synthetic graph as input to various state-of-the-art GNN architectures. Experimental results on three real-world datasets show that HoGS significantly outperforms baseline methods in the accuracy of training GNNs.