Gated Graph Sequence Neural Networks

Abstract

Graph-structured data appears frequently in domains including chemistry, natural language semantics, social networks, and knowledge bases. In this work, we study feature learning techniques for graph-structured inputs. Our starting point is previous work on Graph Neural Networks (Scarselli et al., 2009), which we modify to use gated recurrent units and modern optimization techniques and then extend to output sequences. The result is a flexible and broadly useful class of neural network models that has favorable inductive biases relative to purely sequence-based models (e.g., LSTMs) when the problem is graph-structured. We demonstrate the capabilities on some simple AI (bAbI) and graph algorithm learning tasks. We then show it achieves state-of-the-art performance on a problem from program verification, in which subgraphs need to be matched to abstract data structures.

Published as a conference paper in ICLR 2016. Fixed a typo

References in corpus (4)

• Adam: A Method for Stochastic Optimization
• Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
• Convolutional Networks on Graphs for Learning Molecular Fingerprints
• Ask Me Anything: Dynamic Memory Networks for Natural Language Processing

Cited by in corpus (227)

• Geometric deep learning: going beyond Euclidean data
• Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals
• Junction Tree Variational Autoencoder for Molecular Graph Generation
• Few-Shot Learning with Graph Neural Networks
• A Generalization of Transformer Networks to Graphs
• Link Prediction Based on Graph Neural Networks
• Graph networks as learnable physics engines for inference and control
• Fi-GNN: Modeling Feature Interactions via Graph Neural Networks for CTR Prediction
• DeepGCNs: Making GCNs Go as Deep as CNNs
• Metrics for Graph Comparison: A Practitioner's Guide
• Towards Sparse Hierarchical Graph Classifiers
• Learning a SAT Solver from Single-Bit Supervision
• Adversarial Attack on Graph Structured Data
• Structured Neural Summarization
• Flexible Neural Representation for Physics Prediction
• Learning Multimodal Graph-to-Graph Translation for Molecular Optimization
• Understanding Pooling in Graph Neural Networks
• Message-passing neural networks for high-throughput polymer screening
• Hierarchical Generation of Molecular Graphs using Structural Motifs
• DeepGCNs: Can GCNs Go as Deep as CNNs?
• Covariant Compositional Networks For Learning Graphs
• Relation Modeling with Graph Convolutional Networks for Facial Action Unit Detection
• D-VAE: A Variational Autoencoder for Directed Acyclic Graphs
• Towards Deeper Graph Neural Networks with Differentiable Group Normalization
• Deeply learning molecular structure-property relationships using attention- and gate-augmented graph convolutional network
• Learning Graph-Level Representation for Drug Discovery
• A Generalization of Convolutional Neural Networks to Graph-Structured Data
• Learning Graph Representations with Embedding Propagation
• Neural Networks and Quantum Field Theory
• Compositional Generalization in Semantic Parsing: Pre-training vs. Specialized Architectures
• Learning to Represent Edits
• Contrastive Learning of Structured World Models
• Graph Neural Networks for Natural Language Processing: A Survey
• Spatio-Temporal Neural Networks for Space-Time Series Forecasting and Relations Discovery
• SOLVER: Scene-Object Interrelated Visual Emotion Reasoning Network
• Multiple-Human Parsing in the Wild
• A Two-Step Graph Convolutional Decoder for Molecule Generation
• Context-Dependent Diffusion Network for Visual Relationship Detection
• Large Scale Graph Learning from Smooth Signals
• Are Powerful Graph Neural Nets Necessary? A Dissection on Graph Classification
• Graph-MLP: Node Classification without Message Passing in Graph
• Structured Sequence Modeling with Graph Convolutional Recurrent Networks
• Graph Structure Learning for Robust Graph Neural Networks
• Hyperbolic Attention Networks
• EEG-Based Emotion Recognition Using Regularized Graph Neural Networks
• Graph Convolutional Neural Networks with Node Transition Probability-based Message Passing and DropNode Regularization
• AEGCN: An Autoencoder-Constrained Graph Convolutional Network
• Learning the Travelling Salesperson Problem Requires Rethinking Generalization
• All SMILES Variational Autoencoder
• Graph Normalizing Flows
• Dirichlet Energy Constrained Learning for Deep Graph Neural Networks
• Learning to Extend Molecular Scaffolds with Structural Motifs
• Spatio-Temporal Graph Transformer Networks for Pedestrian Trajectory Prediction
• Natural Graph Networks
• Detecting Code Clones with Graph Neural Networkand Flow-Augmented Abstract Syntax Tree
• Neural Graph Evolution: Towards Efficient Automatic Robot Design
• Graph Star Net for Generalized Multi-Task Learning
• M-Evolve: Structural-Mapping-Based Data Augmentation for Graph Classification
• Inductive Matrix Completion Based on Graph Neural Networks
• Spatio-Temporal Graph Convolution for Skeleton Based Action Recognition
• Edge-Featured Graph Attention Network
• Learning Semantic-Specific Graph Representation for Multi-Label Image Recognition
• AMPNet: Asynchronous Model-Parallel Training for Dynamic Neural Networks
• Hierarchical Graph-to-Graph Translation for Molecules
• Premise Selection for Theorem Proving by Deep Graph Embedding
• NISER: Normalized Item and Session Representations to Handle Popularity Bias
• Learning to map source code to software vulnerability using code-as-a-graph
• Iterative Visual Reasoning Beyond Convolutions
• Learning Efficient Algorithms with Hierarchical Attentive Memory
• Auto-Encoding Scene Graphs for Image Captioning
• Graph Convolution: A High-Order and Adaptive Approach
• Graph Neural Networks for Leveraging Industrial Equipment Structure: An application to Remaining Useful Life Estimation
• GraphFlow: Exploiting Conversation Flow with Graph Neural Networks for Conversational Machine Comprehension
• What graph neural networks cannot learn: depth vs width
• Deriving Neural Architectures from Sequence and Graph Kernels
• Embedding API Dependency Graph for Neural Code Generation
• GraphCrop: Subgraph Cropping for Graph Classification
• Distilling Knowledge from Graph Convolutional Networks
• Adversarial Defense Framework for Graph Neural Network
• Gaussian Attention Model and Its Application to Knowledge Base Embedding and Question Answering
• Road Scene Graph: A Semantic Graph-Based Scene Representation Dataset for Intelligent Vehicles
• Residual or Gate? Towards Deeper Graph Neural Networks for Inductive Graph Representation Learning
• Graph Edge Convolutional Neural Networks for Skeleton Based Action Recognition
• Software Vulnerability Detection via Deep Learning over Disaggregated Code Graph Representation
• Strong Generalization and Efficiency in Neural Programs
• HONEM: Learning Embedding for Higher Order Networks
• Bridging Knowledge Graphs to Generate Scene Graphs
• SPI: Automated Identification of Security Patches via Commits
• Bayesian Semi-supervised Learning with Graph Gaussian Processes
• PAN: Path Integral Based Convolution for Deep Graph Neural Networks
• How hard is to distinguish graphs with graph neural networks?
• Learning Blended, Precise Semantic Program Embeddings
• Learning Scalable and Precise Representation of Program Semantics
• Evolutionary Architecture Search for Graph Neural Networks
• Balancing Multi-level Interactions for Session-based Recommendation
• Evolving Graphical Planner: Contextual Global Planning for Vision-and-Language Navigation
• Sampling methods for efficient training of graph convolutional networks: A survey
• Visual Relationship Detection using Scene Graphs: A Survey
• Natural Question Generation with Reinforcement Learning Based Graph-to-Sequence Model
• A PAC-Bayesian Approach to Generalization Bounds for Graph Neural Networks
• NODE-SELECT: A Graph Neural Network Based On A Selective Propagation Technique
• Graph Deconvolutional Generation
• Graph Neural Networks for the Prediction of Substrate-Specific Organic Reaction Conditions
• Focusing and Diffusion: Bidirectional Attentive Graph Convolutional Networks for Skeleton-based Action Recognition
• Multi-dimensional Graph Convolutional Networks
• Factor Graph Neural Network
• CasGCN: Predicting future cascade growth based on information diffusion graph
• Semi-supervised learning of hierarchical representations of molecules using neural message passing
• Plan, Attend, Generate: Planning for Sequence-to-Sequence Models
• Using Graph Neural Networks for Mass Spectrometry Prediction
• Recursive Graphical Neural Networks for Text Classification
• Benchmarks for Graph Embedding Evaluation
• Towards Interpretable Sparse Graph Representation Learning with Laplacian Pooling
• Learning to Execute Programs with Instruction Pointer Attention Graph Neural Networks
• DyGCN: Dynamic Graph Embedding with Graph Convolutional Network
• Structural Neural Encoders for AMR-to-text Generation
• Learning Visual Commonsense for Robust Scene Graph Generation
• Knowledge-Embedded Representation Learning for Fine-Grained Image Recognition
• A Review on Intelligent Object Perception Methods Combining Knowledge-based Reasoning and Machine Learning
• You are AllSet: A Multiset Function Framework for Hypergraph Neural Networks
• Pileup mitigation at the Large Hadron Collider with Graph Neural Networks
• Revisiting Graph Convolutional Network on Semi-Supervised Node Classification from an Optimization Perspective
• Learning Graph-Level Representations with Recurrent Neural Networks
• Graph Convolutional Networks with EigenPooling
• Learning a Static Bug Finder from Data
• Self-Enhanced GNN: Improving Graph Neural Networks Using Model Outputs
• Artificial Intelligence in Drug Discovery: Applications and Techniques
• SEED: Semantic Graph based Deep detection for type-4 clone
• Jointly Modeling Aspect and Sentiment with Dynamic Heterogeneous Graph Neural Networks
• Face-GCN: A Graph Convolutional Network for 3D Dynamic Face Identification/Recognition
• DAGCN: Dual Attention Graph Convolutional Networks
• Search for emerging jets in $pp$ collisions at $\sqrt{s}=13.6$ TeV with the ATLAS experiment
• DGTN: Dual-channel Graph Transition Network for Session-based Recommendation
• HyperSAGE: Generalizing Inductive Representation Learning on Hypergraphs
• Fusion Graph Convolutional Networks
• Semantic Relationships Guided Representation Learning for Facial Action Unit Recognition
• Amortized Bethe Free Energy Minimization for Learning MRFs
• Graph Signal Processing -- Part III: Machine Learning on Graphs, from Graph Topology to Applications
• User Memory Reasoning for Conversational Recommendation
• Topological based classification using graph convolutional networks
• Generalized Zero-shot ICD Coding
• Variable Name Recovery in Decompiled Binary Code using Constrained Masked Language Modeling
• Modeling Attention Flow on Graphs
• Look Again at the Syntax: Relational Graph Convolutional Network for Gendered Ambiguous Pronoun Resolution
• Neural Message Passing for Multi-Label Classification
• MMGCN: Multimodal Fusion via Deep Graph Convolution Network for Emotion Recognition in Conversation
• A pipeline for fair comparison of graph neural networks in node classification tasks
• EvoNet: A Neural Network for Predicting the Evolution of Dynamic Graphs
• An Overview on the Application of Graph Neural Networks in Wireless Networks
• DAGNN: Demand-aware Graph Neural Networks for Session-based Recommendation
• Graph Neural Network Architecture Search for Molecular Property Prediction
• Demand Forecasting from Spatiotemporal Data with Graph Networks and Temporal-Guided Embedding
• Pose-GNN : Camera Pose Estimation System Using Graph Neural Networks
• Next Waves in Veridical Network Embedding
• Generalization bounds for graph convolutional neural networks via Rademacher complexity
• Equivariant Entity-Relationship Networks
• Train on Small, Play the Large: Scaling Up Board Games with AlphaZero and GNN
• Lymph Node Gross Tumor Volume Detection in Oncology Imaging via Relationship Learning Using Graph Neural Network
• Tree Structure-Aware Graph Representation Learning via Integrated Hierarchical Aggregation and Relational Metric Learning
• Collaborative Motion Prediction via Neural Motion Message Passing
• GraphSearchNet: Enhancing GNNs via Capturing Global Dependencies for Semantic Code Search
• CommPOOL: An Interpretable Graph Pooling Framework for Hierarchical Graph Representation Learning
• Adversarial Graph Representation Adaptation for Cross-Domain Facial Expression Recognition
• Graph Neural Networks for Image Understanding Based on Multiple Cues: Group Emotion Recognition and Event Recognition as Use Cases
• Enhance Information Propagation for Graph Neural Network by Heterogeneous Aggregations
• CPR-GCN: Conditional Partial-Residual Graph Convolutional Network in Automated Anatomical Labeling of Coronary Arteries
• Graph Neural Networks Including Sparse Interpretability
• Parameter Prediction for Unseen Deep Architectures
• Permutation-Invariant Variational Autoencoder for Graph-Level Representation Learning
• Multi-label Zero-shot Classification by Learning to Transfer from External Knowledge
• Fast and Accurate: Structure Coherence Component for Face Alignment
• Learning Predicates as Functions to Enable Few-shot Scene Graph Prediction
• Drug Package Recommendation via Interaction-aware Graph Induction
• Efficient and Interpretable Robot Manipulation with Graph Neural Networks
• Deep Data Flow Analysis
• Adaptive Neural Message Passing for Inductive Learning on Hypergraphs
• Identifying Non-Control Security-Critical Data through Program Dependence Learning
• Recovering the Unbiased Scene Graphs from the Biased Ones
• Graph Attention Networks with Positional Embeddings
• Deep Dynamic Effective Connectivity Estimation from Multivariate Time Series
• TGG: Transferable Graph Generation for Zero-shot and Few-shot Learning
• Dynamic Graph Generation Network: Generating Relational Knowledge from Diagrams
• Sparse hierarchical representation learning on molecular graphs
• GAIN: Graph Attention & Interaction Network for Inductive Semi-Supervised Learning over Large-scale Graphs
• Out-of-Town Recommendation with Travel Intention Modeling
• Analyzing the Performance of Graph Neural Networks with Pipe Parallelism
• Heterogeneous Graph based Deep Learning for Biomedical Network Link Prediction
• Improving the Long-Range Performance of Gated Graph Neural Networks
• Asking Complex Questions with Multi-hop Answer-focused Reasoning
• Convergent Graph Solvers
• A Context Integrated Relational Spatio-Temporal Model for Demand and Supply Forecasting
• Fea2Fea: Exploring Structural Feature Correlations via Graph Neural Networks
• Framework for Designing Filters of Spectral Graph Convolutional Neural Networks in the Context of Regularization Theory
• Neural Topic Modeling by Incorporating Document Relationship Graph
• Fast Mesh Data Augmentation via Chebyshev Polynomial of Spectral filtering
• Learning Robust Representations with Graph Denoising Policy Network
• Estimating Fund-Raising Performance for Start-up Projects from a Market Graph Perspective
• Critical Percolation as a Framework to Analyze the Training of Deep Networks
• Towards explainable message passing networks for predicting carbon dioxide adsorption in metal-organic frameworks
• Lorentzian Graph Convolutional Networks
• ShadowGNN: Graph Projection Neural Network for Text-to-SQL Parser
• Which Hyperparameters to Optimise? An Investigation of Evolutionary Hyperparameter Optimisation in Graph Neural Network For Molecular Property Prediction
• Genetic Constrained Graph Variational Autoencoder for COVID-19 Drug Discovery
• GAHNE: Graph-Aggregated Heterogeneous Network Embedding
• Fixed Priority Global Scheduling from a Deep Learning Perspective
• Are My Invariants Valid? A Learning Approach
• Batched Sparse Matrix Multiplication for Accelerating Graph Convolutional Networks
• On the Significance of Question Encoder Sequence Model in the Out-of-Distribution Performance in Visual Question Answering
• Policy Message Passing: A New Algorithm for Probabilistic Graph Inference
• Keywords Guided Method Name Generation
• Diversified Multiscale Graph Learning with Graph Self-Correction
• Intrinsic Geometric Information Transfer Learning on Multiple Graph-Structured Datasets
• Knowledge as Invariance -- History and Perspectives of Knowledge-augmented Machine Learning
• An Evalutation of Programming Language Models' performance on Software Defect Detection
• MetAL: Active Semi-Supervised Learning on Graphs via Meta Learning
• Graph Representation Learning Network via Adaptive Sampling
• Edge-Labeling based Directed Gated Graph Network for Few-shot Learning
• FeatureNorm: L2 Feature Normalization for Dynamic Graph Embedding
• Understanding in Artificial Intelligence
• Improving Automatic Source Code Summarization via Deep Reinforcement Learning
• Action Recognition with Kernel-based Graph Convolutional Networks
• Estimating Early Fundraising Performance of Innovations via Graph-based Market Environment Model
• DeepDrawing: A Deep Learning Approach to Graph Drawing
• Joint Forecasting and Interpolation of Graph Signals Using Deep Learning
• Graph Feature Gating Networks
• A Method to Predict Semantic Relations on Artificial Intelligence Papers
• Edge-featured Graph Neural Architecture Search