A survey of dynamic graph neural networks

Yanping Zheng,Lu Yi,Zhewei Wei
2024-04-28
Abstract:Graph neural networks (GNNs) have emerged as a powerful tool for effectively mining and learning from graph-structured data, with applications spanning numerous domains. However, most research focuses on static graphs, neglecting the dynamic nature of real-world networks where topologies and attributes evolve over time. By integrating sequence modeling modules into traditional GNN architectures, dynamic GNNs aim to bridge this gap, capturing the inherent temporal dependencies of dynamic graphs for a more authentic depiction of complex networks. This paper provides a comprehensive review of the fundamental concepts, key techniques, and state-of-the-art dynamic GNN models. We present the mainstream dynamic GNN models in detail and categorize models based on how temporal information is incorporated. We also discuss large-scale dynamic GNNs and pre-training techniques. Although dynamic GNNs have shown superior performance, challenges remain in scalability, handling heterogeneous information, and lack of diverse graph datasets. The paper also discusses possible future directions, such as adaptive and memory-enhanced models, inductive learning, and theoretical analysis.
Machine Learning,Social and Information Networks
What problem does this paper attempt to address?
### Problems the Paper Aims to Solve This survey paper, "A Survey of Dynamic Graph Neural Networks," aims to address the following issues: 1. **Limitations of Static Graph Neural Networks**: - Most existing Graph Neural Network (GNN) research focuses on static graphs, neglecting the dynamic nature of real-world networks. In these dynamic graphs, the topology and attributes of nodes and edges change over time. - Static GNNs are difficult to apply directly to dynamic graphs, posing new challenges for the analysis and learning of dynamic graphs. 2. **Development of Dynamic Graph Neural Networks**: - By integrating sequence modeling modules into traditional GNN architectures, dynamic GNNs aim to bridge this gap, capturing the inherent temporal dependencies in dynamic graphs to more accurately depict complex networks. - Dynamic GNNs have shown superior performance but still face challenges in scalability, handling heterogeneous information, and the lack of diverse graph datasets. 3. **Shortcomings of Existing Surveys**: - Although some work has reviewed methods for dynamic graph representation learning, the content of existing surveys has become outdated with the continuous emergence of new methods and applications, failing to reflect the latest research findings and technological trends. - Therefore, this paper provides an up-to-date survey of dynamic GNN models, covering techniques for representing and modeling dynamic graphs and comparing existing models. 4. **Future Research Directions**: - Possible future research directions are explored, such as adaptive and enhanced memory models, inductive learning, and theoretical analysis. ### Main Contributions - **Comprehensive Survey**: This paper provides a comprehensive survey of the basic concepts and characteristics of dynamic GNNs, constructing a knowledge system for dynamic GNNs and systematically investigating key technologies and models. - **Detailed Analysis**: It introduces current mainstream dynamic GNN models in detail and deeply analyzes their advantages and disadvantages. - **Challenges and Future Directions**: The challenges in this field are discussed, highlighting potential future research directions and technological trends. ### Structure Overview - **Part 1**: Introduction, introducing the importance of graph neural networks in handling complex graph-structured data and the advantages of dynamic graphs over static graphs. - **Part 2**: Symbols and background knowledge, defining relevant terms and mathematical representations of dynamic graphs. - **Part 3**: Problem statement of dynamic graph representation learning, comprehensively reviewing dynamic GNNs according to the classification shown in Figure 1. - **Part 4**: Models and systems for large-scale dynamic graphs. - **Part 5**: Commonly used datasets, prediction tasks, and benchmarks. - **Part 6**: Introduction of new techniques recently applied to dynamic graph learning, such as transfer learning and pre-training techniques. - **Part 7**: Discussion of possible future research directions. - **Part 8**: Conclusion. ### Summary By comprehensively reviewing the latest developments in dynamic GNNs, this paper aims to fill the gaps in existing surveys, providing researchers with a systematic reference framework to promote further development in this field.