Graph neural networks: a survey on the links between privacy and security
Графовые нейронные сети: обзор взаимосвязей между конфиденциальностью и безопасностью
2024-02-08
SCID: 54.1/ak9yz69f
Discuss with AI
GNN privacyGNN securitygraph adversarial attacksgraph neural networksmodel stealing
Figures from the paper
Abstract (AI)
Abstract Graph neural networks (GNNs) are models that capture the dependencies between graph data by passing messages between graph nodes and they have been widely used to process graph data that contains relational information. Example application areas include social networks, recommendation systems, and life sciences. However, like all neural networks, there are underpinning security and privacy concerns associated with GNN deployments in practice. For example, attackers can perturb a graph’s data to undermine a model’s effectiveness, or they can steal the model’s data and/or parameters, thus threatening the privacy of the model. In this survey, we provide a comprehensive review of recent research efforts on security and/or privacy in GNNs. We also systematically describe the distinctions and relationships between security and privacy, as well as providing an outlook on future directions of research in this area.
Key Findings
1
GNN deployments face privacy threats involving theft of model data and/or parameters.
2
GNN deployments face security threats in which adversaries perturb graph data to reduce model effectiveness.
3
Graph neural networks process relational graph data through message passing and are widely applied in social networks, recommendation systems, and life sciences.
4
The survey comprehensively reviews research on GNN security and privacy, clarifies distinctions and relationships between these concepts, and outlines future research directions.
Research Object
graph neural network deployments processing relational graph data
Research Subject
security and privacy vulnerabilities, distinctions, and interrelationships in graph neural networks
Publication Details
Publication Date
2024-02-08
Journal
Publisher
ISSN
Cited by
42
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai12
Emergence of Scaling in Random Networks1999
Semi-Supervised Classification with Graph Convolutional Networks2016
Distilling the Knowledge in a Neural Network2015
Machine learning: Trends, perspectives, and prospects2015
The Graph Neural Network Model2008
Graph neural networks: A review of methods and applications2020
Deep Learning with Differential Privacy2016
Graph Convolutional Neural Networks for Web-Scale Recommender Systems2018
Heterogeneous Graph Attention Network2019
Graph Neural Networks in Recommender Systems: A Survey2022
MAGNN: Metapath Aggregated Graph Neural Network for Heterogeneous Graph Embedding2020
Graph Neural Networks for Natural Language Processing: A Survey2023