Representation Learning for Dynamic Graphs: A Survey

Abstract

Graphs arise naturally in many real-world applications including social networks, recommender systems, ontologies, biology, and computational finance. Traditionally, machine learning models for graphs have been mostly designed for static graphs. However, many applications involve evolving graphs. This introduces important challenges for learning and inference since nodes, attributes, and edges change over time. In this survey, we review the recent advances in representation learning for dynamic graphs, including dynamic knowledge graphs. We describe existing models from an encoder-decoder perspective, categorize these encoders and decoders based on the techniques they employ, and analyze the approaches in each category. We also review several prominent applications and widely used datasets and highlight directions for future research.

Accepted at JMLR, 73 pages, 2 figures

References in corpus (46)

• Distributed Representations of Words and Phrases and their Compositionality
• Neural Message Passing for Quantum Chemistry
• Geometric deep learning: going beyond Euclidean data
• Teaching Machines to Read and Comprehend
• KGAT: Knowledge Graph Attention Network for Recommendation
• MXNet: A Flexible and Efficient Machine Learning Library for Heterogeneous Distributed Systems
• A Review of Relational Machine Learning for Knowledge Graphs
• Bidirectional Attention Flow for Machine Comprehension
• Graph Neural Networks: A Review of Methods and Applications
• Fast Graph Representation Learning with PyTorch Geometric
• Biological network comparison using graphlet degree distribution
• What's in a crowd? Analysis of face-to-face behavioral networks
• dyngraph2vec: Capturing Network Dynamics using Dynamic Graph Representation Learning
• Dynamic Graph Convolutional Networks
• Go for a Walk and Arrive at the Answer: Reasoning Over Paths in Knowledge Bases using Reinforcement Learning
• Attributed Network Embedding for Learning in a Dynamic Environment
• Community Interaction and Conflict on the Web
• DynGEM: Deep Embedding Method for Dynamic Graphs
• Scalable Link Prediction in Dynamic Networks via Non-Negative Matrix Factorization
• Analogical Inference for Multi-Relational Embeddings
• Hypernetwork Knowledge Graph Embeddings
• LanczosNet: Multi-Scale Deep Graph Convolutional Networks
• Dynamic Word Embeddings
• Node Embedding over Temporal Graphs
• Counting Belief Propagation
• Universal Invariant and Equivariant Graph Neural Networks
• GC-LSTM: Graph Convolution Embedded LSTM for Dynamic Link Prediction
• TEQUILA: Temporal Question Answering over Knowledge Bases
• On the representation and embedding of knowledge bases beyond binary relations
• Efficiently Embedding Dynamic Knowledge Graphs
• Deep Coevolutionary Network: Embedding User and Item Features for Recommendation
• Embedding Logical Queries on Knowledge Graphs
• Neural Embeddings of Graphs in Hyperbolic Space
• Time2Vec: Learning a Vector Representation of Time
• On Multi-Relational Link Prediction with Bilinear Models
• Building Dynamic Knowledge Graphs from Text using Machine Reading Comprehension
• 3D Graph Convolutional Networks with Temporal Graphs: A Spatial Information Free Framework For Traffic Forecasting
• Automorphism Groups of Graphical Models and Lifted Variational Inference
• Temporal Knowledge Graph Embedding Model based on Additive Time Series Decomposition
• On the Equivalence of Holographic and Complex Embeddings for Link Prediction
• Efficient Representation Learning Using Random Walks for Dynamic Graphs
• STWalk: Learning Trajectory Representations in Temporal Graphs
• New Liftable Classes for First-Order Probabilistic Inference
• RelNN: A Deep Neural Model for Relational Learning
• Abstractive Summarization Using Attentive Neural Techniques
• DyANE: Dynamics-aware node embedding for temporal networks

Cited by in corpus (9)

• Anomaly Detection in Dynamic Graphs: A Comprehensive Survey
• Efficient Scaling of Dynamic Graph Neural Networks
• Graph Deep Learning for Time Series Forecasting
• A Self-supervised Riemannian GNN with Time Varying Curvature for Temporal Graph Learning
• TEAM: Topological Evolution-aware Framework for Traffic Forecasting--Extended Version
• Inductive Graph Transformer for Delivery Time Estimation
• Dynamic Graph Unlearning: A General and Efficient Post-Processing Method via Gradient Transformation
• TIDFormer: Exploiting Temporal and Interactive Dynamics Makes A Great Dynamic Graph Transformer
• CoCon: A Data Set on Combined Contextualized Research Artifact Use