GNNExplainer: Generating Explanations for Graph Neural Networks

Abstract

Graph Neural Networks (GNNs) are a powerful tool for machine learning on graphs.GNNs combine node feature information with the graph structure by recursively passing neural messages along edges of the input graph. However, incorporating both graph structure and feature information leads to complex models, and explaining predictions made by GNNs remains unsolved. Here we propose GNNExplainer, the first general, model-agnostic approach for providing interpretable explanations for predictions of any GNN-based model on any graph-based machine learning task. Given an instance, GNNExplainer identifies a compact subgraph structure and a small subset of node features that have a crucial role in GNN's prediction. Further, GNNExplainer can generate consistent and concise explanations for an entire class of instances. We formulate GNNExplainer as an optimization task that maximizes the mutual information between a GNN's prediction and distribution of possible subgraph structures. Experiments on synthetic and real-world graphs show that our approach can identify important graph structures as well as node features, and outperforms baselines by 17.1% on average. GNNExplainer provides a variety of benefits, from the ability to visualize semantically relevant structures to interpretability, to giving insights into errors of faulty GNNs.

References in corpus (21)

• Inductive Representation Learning on Large Graphs
• Neural Message Passing for Quantum Chemistry
• Towards A Rigorous Science of Interpretable Machine Learning
• Axiomatic Attribution for Deep Networks
• Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
• Learning Important Features Through Propagating Activation Differences
• Modeling polypharmacy side effects with graph convolutional networks
• Graph Neural Networks: A Review of Methods and Applications
• Understanding Black-box Predictions via Influence Functions
• Representation Learning on Graphs with Jumping Knowledge Networks
• Sanity Checks for Saliency Maps
• Discriminative Embeddings of Latent Variable Models for Structured Data
• Visualizing Deep Neural Network Decisions: Prediction Difference Analysis
• Community Interaction and Conflict on the Web
• Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation
• Adaptive Sampling Towards Fast Graph Representation Learning
• Predicting Organic Reaction Outcomes with Weisfeiler-Lehman Network
• Deep Learning on Graphs: A Survey
• Representer Point Selection for Explaining Deep Neural Networks
• Interpretable Graph Convolutional Neural Networks for Inference on Noisy Knowledge Graphs
• ExplaiNE: An Approach for Explaining Network Embedding-based Link Predictions

Cited by in corpus (72)

• Explaining Deep Neural Networks and Beyond: A Review of Methods and Applications
• Graph Neural Network for Traffic Forecasting: A Survey
• Recent Advances and Applications of Deep Learning Methods in Materials Science
• Explainable Artificial Intelligence Applications in Cyber Security: State-of-the-Art in Research
• E-GraphSAGE: A Graph Neural Network based Intrusion Detection System for IoT
• Graph-based Deep Learning for Communication Networks: A Survey
• Discovering Symbolic Models from Deep Learning with Inductive Biases
• Graph-Based Deep Learning for Medical Diagnosis and Analysis: Past, Present and Future
• Causal Attention for Interpretable and Generalizable Graph Classification
• Multimodal Data Integration for Oncology in the Era of Deep Neural Networks: A Review
• A Survey on Graph-Based Deep Learning for Computational Histopathology
• Toward Explainable AI for Regression Models
• Advances of Machine Learning in Materials Science: Ideas and Techniques
• A Comprehensive Survey on Trustworthy Graph Neural Networks: Privacy, Robustness, Fairness, and Explainability
• Graph Rationalization with Environment-based Augmentations
• On Proximity and Structural Role-based Embeddings in Networks: Misconceptions, Techniques, and Applications
• Unnoticeable Backdoor Attacks on Graph Neural Networks
• Scientometric analysis and knowledge mapping of literature-based discovery (1986-2020)
• Actionable and Interpretable Fault Localization for Recurring Failures in Online Service Systems
• Should Graph Convolution Trust Neighbors? A Simple Causal Inference Method
• Entropy-based Logic Explanations of Neural Networks
• On the (In)fidelity and Sensitivity for Explanations
• DIG: A Turnkey Library for Diving into Graph Deep Learning Research
• SE-GSL: A General and Effective Graph Structure Learning Framework through Structural Entropy Optimization
• xFraud: Explainable Fraud Transaction Detection
• Toward Accurate Interpretable Predictions of Materials Properties within Transformer Language Models
• A Heterogeneous Graph-Based Multi-Task Learning for Fault Event Diagnosis in Smart Grid
• Predicting Biomedical Interactions with Higher-Order Graph Convolutional Networks
• Interpretable Graph Convolutional Network of Multi-Modality Brain Imaging for Alzheimer's Disease Diagnosis
• Time Series Predictions in Unmonitored Sites: A Survey of Machine Learning Techniques in Water Resources
• Can Graph Neural Networks Count Substructures?
• Understanding the Message Passing in Graph Neural Networks via Power Iteration Clustering
• Implementing graph neural networks with TensorFlow-Keras
• A Meta-Learning Approach for Training Explainable Graph Neural Networks
• MixupExplainer: Generalizing Explanations for Graph Neural Networks with Data Augmentation
• ChemBoost: A chemical language based approach for protein-ligand binding affinity prediction
• Visualizing Graph Neural Networks with CorGIE: Corresponding a Graph to Its Embedding
• A Network Science perspective of Graph Convolutional Networks: A survey
• Projective Ranking-based GNN Evasion Attacks
• View-based Explanations for Graph Neural Networks
• Differentiable graph-structured models for inverse design of lattice materials
• Towards Human-like Perception: Learning Structural Causal Model in Heterogeneous Graph
• Understanding Isomorphism Bias in Graph Data Sets
• GRETEL: A unified framework for Graph Counterfactual Explanation Evaluation
• Goal-directed graph construction using reinforcement learning
• DASH: Dynamic Attention-Based Substructure Hierarchy for Partial Charge Assignment
• Optimal Solving of Constrained Path-Planning Problems with Graph Convolutional Networks and Optimized Tree Search
• Knowledge Graph Embeddings and Explainable AI
• Deep Graph Similarity Learning: A Survey
• X-CBA: Explainability Aided CatBoosted Anomal-E for Intrusion Detection System
• Detection, Explanation and Filtering of Cyber Attacks Combining Symbolic and Sub-Symbolic Methods
• Adversarial Model Extraction on Graph Neural Networks
• Discovering robust biomarkers of psychiatric disorders from resting-state functional MRI via graph neural networks: A systematic review
• DRExplainer: Quantifiable Interpretability in Drug Response Prediction with Directed Graph Convolutional Network
• Finding Heterophilic Neighbors via Confidence-based Subgraph Matching for Semi-supervised Node Classification
• i-Align: an interpretable knowledge graph alignment model
• Auto-Configured Explainable Graph Neural Networks for Multi-Site Pollution Prediction
• Technical Understanding from IML Hands-on Experience: A Study through a Public Event for Science Museum Visitors
• Graph Homomorphism Convolution
• Explain Graph Neural Networks to Understand Weighted Graph Features in Node Classification
• SpotV2Net: Multivariate Intraday Spot Volatility Forecasting via Vol-of-Vol-Informed Graph Attention Networks
• Boosting Column Generation with Graph Neural Networks for Joint Rider Trip Planning and Crew Shift Scheduling
• Explainable Link Prediction for Privacy-Preserving Contact Tracing
• Relational Graph Neural Networks for Fraud Detection in a Super-App environment
• Is Your Explanation Reliable: Confidence-Aware Explanation on Graph Neural Networks
• Mixture-of-Experts Graph Transformers for Interpretable Particle Collision Detection
• Learnt Sparsification for Interpretable Graph Neural Networks
• Explaining machine-learned particle-flow reconstruction
• On the Robustness of Deep Learning-predicted Contention Models for Network Calculus
• SEEN: Sharpening Explanations for Graph Neural Networks using Explanations from Neighborhoods
• Explainability in subgraphs-enhanced Graph Neural Networks
• How Faithful are Self-Explainable GNNs?