Trustworthy Graph Neural Networks: Aspects, Methods and Trends

Abstract

Graph neural networks (GNNs) have emerged as a series of competent graph learning methods for diverse real-world scenarios, ranging from daily applications like recommendation systems and question answering to cutting-edge technologies such as drug discovery in life sciences and n-body simulation in astrophysics. However, task performance is not the only requirement for GNNs. Performance-oriented GNNs have exhibited potential adverse effects like vulnerability to adversarial attacks, unexplainable discrimination against disadvantaged groups, or excessive resource consumption in edge computing environments. To avoid these unintentional harms, it is necessary to build competent GNNs characterised by trustworthiness. To this end, we propose a comprehensive roadmap to build trustworthy GNNs from the view of the various computing technologies involved. In this survey, we introduce basic concepts and comprehensively summarise existing efforts for trustworthy GNNs from six aspects, including robustness, explainability, privacy, fairness, accountability, and environmental well-being. Additionally, we highlight the intricate cross-aspect relations between the above six aspects of trustworthy GNNs. Finally, we present a thorough overview of trending directions for facilitating the research and industrialisation of trustworthy GNNs.

42 pages, 7 tables, 4 figures, double columns, accepted by Proceedings of the IEEE

References in corpus (87)

• A Comprehensive Survey on Graph Neural Networks
• Deep Learning with Differential Privacy
• Federated Learning: Challenges, Methods, and Future Directions
• Graph Convolutional Neural Networks for Web-Scale Recommender Systems
• Federated Learning with Non-IID Data
• Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks
• Deep Graph Library: A Graph-Centric, Highly-Performant Package for Graph Neural Networks
• Enhancing Graph Neural Network-based Fraud Detectors against Camouflaged Fraudsters
• Graph Self-Supervised Learning: A Survey
• Causal Intervention for Leveraging Popularity Bias in Recommendation
• Learning Graph Embedding with Adversarial Training Methods
• Adversarial Attacks on Graph Neural Networks via Meta Learning
• A Survey on the Fairness of Recommender Systems
• XGNN: Towards Model-Level Explanations of Graph Neural Networks
• Higher-Order Explanations of Graph Neural Networks via Relevant Walks
• Slalom: Fast, Verifiable and Private Execution of Neural Networks in Trusted Hardware
• Toward Trustworthy AI Development: Mechanisms for Supporting Verifiable Claims
• ETA Prediction with Graph Neural Networks in Google Maps
• Towards Long-term Fairness in Recommendation
• Federated Social Recommendation with Graph Neural Network
• Transferring Robustness for Graph Neural Network Against Poisoning Attacks
• DarkneTZ: Towards Model Privacy at the Edge using Trusted Execution Environments
• Fairness Under Unawareness: Assessing Disparity When Protected Class Is Unobserved
• Fast Gradient Attack on Network Embedding
• EDITS: Modeling and Mitigating Data Bias for Graph Neural Networks
• Graph Vulnerability and Robustness: A Survey
• Anomaly Detection in Dynamic Graphs via Transformer
• OGB-LSC: A Large-Scale Challenge for Machine Learning on Graphs
• PinnerSage: Multi-Modal User Embedding Framework for Recommendations at Pinterest
• An Empirical Survey on Long Document Summarization: Datasets, Models and Metrics
• Robust Explainability: A Tutorial on Gradient-Based Attribution Methods for Deep Neural Networks
• GraphVite: A High-Performance CPU-GPU Hybrid System for Node Embedding
• Graph Neural Networks for Graphs with Heterophily: A Survey
• Certifiable Robustness and Robust Training for Graph Convolutional Networks
• FedGraphNN: A Federated Learning System and Benchmark for Graph Neural Networks
• Deep Learning for Android Malware Defenses: a Systematic Literature Review
• Learning and Evaluating Graph Neural Network Explanations based on Counterfactual and Factual Reasoning
• IoT-based Android Malware Detection Using Graph Neural Network With Adversarial Defense
• FairDrop: Biased Edge Dropout for Enhancing Fairness in Graph Representation Learning
• Explainability in Graph Neural Networks: A Taxonomic Survey
• Quantifying Privacy Leakage in Graph Embedding
• TDGIA:Effective Injection Attacks on Graph Neural Networks
• Differentially Private Federated Knowledge Graphs Embedding
• Learning Fair Node Representations with Graph Counterfactual Fairness
• NF-GNN: Network Flow Graph Neural Networks for Malware Detection and Classification
• Subgraph Federated Learning with Missing Neighbor Generation
• Accurate, Efficient and Scalable Graph Embedding
• Transformer for Graphs: An Overview from Architecture Perspective
• Automated Machine Learning on Graphs: A Survey
• Federated Graph Classification over Non-IID Graphs
• Accelerating Large Scale Real-Time GNN Inference using Channel Pruning
• Node-Level Membership Inference Attacks Against Graph Neural Networks
• FairSR: Fairness-aware Sequential Recommendation through Multi-Task Learning with Preference Graph Embeddings
• xFraud: Explainable Fraud Transaction Detection
• Learning Strong Graph Neural Networks with Weak Information
• RawlsGCN: Towards Rawlsian Difference Principle on Graph Convolutional Network
• GraphFL: A Federated Learning Framework for Semi-Supervised Node Classification on Graphs
• Attending to Graph Transformers
• Interpretable Graph Convolutional Network of Multi-Modality Brain Imaging for Alzheimer's Disease Diagnosis
• SpreadGNN: Serverless Multi-task Federated Learning for Graph Neural Networks
• Bag of Tricks for Node Classification with Graph Neural Networks
• A Vertical Federated Learning Framework for Graph Convolutional Network
• Out-Of-Distribution Generalization on Graphs: A Survey
• NetFense: Adversarial Defenses against Privacy Attacks on Neural Networks for Graph Data
• GraphSAC: Detecting anomalies in large-scale graphs
• Federated Graph Learning -- A Position Paper
• Vertically Federated Graph Neural Network for Privacy-Preserving Node Classification
• Probing GNN Explainers: A Rigorous Theoretical and Empirical Analysis of GNN Explanation Methods
• DefenseVGAE: Defending against Adversarial Attacks on Graph Data via a Variational Graph Autoencoder
• Adversarial Inter-Group Link Injection Degrades the Fairness of Graph Neural Networks
• Learned Low Precision Graph Neural Networks
• FedGCN: Convergence-Communication Tradeoffs in Federated Training of Graph Convolutional Networks
• Unraveling Privacy Risks of Individual Fairness in Graph Neural Networks
• A Survey on Graph Neural Network Acceleration: Algorithms, Systems, and Customized Hardware
• STFL: A Temporal-Spatial Federated Learning Framework for Graph Neural Networks
• A Novel Genetic Algorithm with Hierarchical Evaluation Strategy for Hyperparameter Optimisation of Graph Neural Networks
• Study of Mass Composition of Cosmic Rays with IceTop and IceCube
• Structure-free Graph Condensation: From Large-scale Graphs to Condensed Graph-free Data
• Towards Data-centric Graph Machine Learning: Review and Outlook
• Membership Inference Attacks on Knowledge Graphs
• Adversarial Immunization for Certifiable Robustness on Graphs
• Towards Representation Identical Privacy-Preserving Graph Neural Network via Split Learning
• Jointly Attacking Graph Neural Network and its Explanations
• A pipeline for fair comparison of graph neural networks in node classification tasks
• Towards the Explanation of Graph Neural Networks in Digital Pathology with Information Flows
• Holistic Survey of Privacy and Fairness in Machine Learning
• Lightning: Striking the Secure Isolation on GPU Clouds with Transient Hardware Faults

Cited by in corpus (12)

• Unifying Large Language Models and Knowledge Graphs: A Roadmap
• A Survey of Graph Neural Networks for Social Recommender Systems
• A Comprehensive Survey on Trustworthy Graph Neural Networks: Privacy, Robustness, Fairness, and Explainability
• Beyond Low-pass Filtering: Graph Convolutional Networks with Automatic Filtering
• TrustGuard: GNN-based Robust and Explainable Trust Evaluation with Dynamicity Support
• Towards Graph Self-Supervised Learning with Contrastive Adjusted Zooming
• Counterfactual Learning on Graphs: A Survey
• Analysis of Gene Regulatory Networks from Gene Expression Using Graph Neural Networks
• MAPPING: Debiasing Graph Neural Networks for Fair Node Classification with Limited Sensitive Information Leakage
• Toward Fair Graph Neural Networks Via Dual-Teacher Knowledge Distillation
• Community detection robustness of graph neural networks
• Learning-based Privacy-Preserving Graph Publishing Against Sensitive Link Inference Attacks