Multi-View Graph Feature Propagation for Privacy Preservation and Feature Sparsity

2025-10-13

AI Summary

Graph Neural Networks (GNNs) for node classification rely heavily on complete node features; however, their performance degrades significantly under highly sparse or privacy-sensitive feature settings, while also risking privacy leakage. To address this, we propose the Multi-View Feature Propagation (MVFP) framework: it partitions raw features into multiple disjoint subviews, injects independent Gaussian noise into each, and performs graph-based propagation separately—thereby preventing reconstruction of original features and jointly optimizing utility and privacy. Our key innovations are: (i) enhancing representation robustness for sparse features via noisy multi-view propagation; and (ii) replacing feature recovery with functional replacement, fundamentally mitigating sensitive information inversion. Extensive experiments on benchmark datasets demonstrate that MVFP achieves substantial gains in classification accuracy over state-of-the-art methods, while reducing membership inference attack success rates by up to 42.6%, validating its effectiveness, stability, and practicality.

Technology Category

• Machine Learning: Privacy
• Natural Language Processing: Safety and Robustness
• Computer Vision: Multi-modal Vision

Application Category

• Security and Privacy: Data transparency and provenance
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems

Abstract

Graph Neural Networks (GNNs) have demonstrated remarkable success in node classification tasks over relational data, yet their effectiveness often depends on the availability of complete node features. In many real-world scenarios, however, feature matrices are highly sparse or contain sensitive information, leading to degraded performance and increased privacy risks. Furthermore, direct exposure of information can result in unintended data leakage, enabling adversaries to infer sensitive information. To address these challenges, we propose a novel Multi-view Feature Propagation (MFP) framework that enhances node classification under feature sparsity while promoting privacy preservation. MFP extends traditional Feature Propagation (FP) by dividing the available features into multiple Gaussian-noised views, each propagating information independently through the graph topology. The aggregated representations yield expressive and robust node embeddings. This framework is novel in two respects: it introduces a mechanism that improves robustness under extreme sparsity, and it provides a principled way to balance utility with privacy. Extensive experiments conducted on graph datasets demonstrate that MFP outperforms state-of-the-art baselines in node classification while substantially reducing privacy leakage. Moreover, our analysis demonstrates that propagated outputs serve as alternative imputations rather than reconstructions of the original features, preserving utility without compromising privacy. A comprehensive sensitivity analysis further confirms the stability and practical applicability of MFP across diverse scenarios. Overall, MFP provides an effective and privacy-aware framework for graph learning in domains characterized by missing or sensitive features.

Problem

• Enhancing node classification under sparse features
• Reducing privacy risks from sensitive graph data
• Balancing model utility with privacy preservation

Innovation

• Multi-view Feature Propagation with Gaussian-noised views
• Independent feature propagation through graph topology
• Balancing utility and privacy in sparse features