Abstract:Dynamic Graph Neural Networks (GNNs) combine temporal information with GNNs to capture structural, temporal, and contextual relationships in dynamic graphs simultaneously, leading to enhanced performance in various applications. As the demand for dynamic GNNs continues to grow, numerous models and frameworks have emerged to cater to different application needs. There is a pressing need for a comprehensive survey that evaluates the performance, strengths, and limitations of various approaches in this domain. This paper aims to fill this gap by offering a thorough comparative analysis and experimental evaluation of dynamic GNNs. It covers 81 dynamic GNN models with a novel taxonomy, 12 dynamic GNN training frameworks, and commonly used benchmarks. We also conduct experimental results from testing representative nine dynamic GNN models and three frameworks on six standard graph datasets. Evaluation metrics focus on convergence accuracy, training efficiency, and GPU memory usage, enabling a thorough comparison of performance across various models and frameworks. From the analysis and evaluation results, we identify key challenges and offer principles for future research to enhance the design of models and frameworks in the dynamic GNNs field.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is some limitations in the current research reviews on Dynamic Graph Neural Networks (DGNNs). Specifically, these limitations include: 1. **Incomplete research coverage** (L1): Most of the existing review literatures are rather outdated and do not cover the DGNN models that have emerged in recent years. 2. **Lack of discussion on DGNN frameworks** (L2): Most review literatures mainly focus on the models themselves and overlook the importance of the frameworks. 3. **Ignoring evaluation benchmarks** (L3): In the existing literatures, the benchmark datasets and metrics used to evaluate DGNN models are rarely introduced in detail. 4. **Lack of experimental comparison** (L4): Although the existing review literatures provide an overview of the models, they lack experimental comparison among different models, which makes it difficult to fully understand the performance differences of each model. 5. **Failure to address emerging application requirements and new challenges** (L5): With the development of technology, DGNN models have great potential in many emerging fields, but the existing research has not fully explored the requirements and challenges in these fields. To address the above problems, the goal of this paper is to provide a comprehensive and up - to - date DGNN review, covering the following main contents: - **Comprehensive model survey and novel classification method** (C1): An extensive survey of 81 recent DGNN models has been carried out, and a new classification method has been proposed to better understand and compare these models. - **Overview of existing DGNN frameworks** (C2): 12 existing DGNN frameworks are introduced in detail, and their characteristics and optimization methods are explored. - **Introduction of DGNN evaluation benchmarks** (C3): Commonly used datasets and evaluation metrics are provided to promote more comprehensive and reproducible experiments. - **Experimental comparison of selected works** (C4): Under a consistent experimental setup, detailed experimental comparisons are carried out on 9 representative DGNN models and 3 frameworks, and the training accuracy, efficiency and memory usage are evaluated. - **Analysis of DGNN challenges** (C5): The new challenges faced by DGNN models and frameworks are explored, and the directions for future research are proposed. Through these contributions, this paper aims to fill the gaps in the existing review literatures and provide a comprehensive and systematic reference resource for researchers and practitioners.

A Comprehensive Survey of Dynamic Graph Neural Networks: Models, Frameworks, Benchmarks, Experiments and Challenges

A survey of dynamic graph neural networks

A Survey on Dynamic Graph Neural Networks Modeling

Dynamic Graph Representation Learning With Neural Networks: A Survey

Foundations and modelling of dynamic networks using Dynamic Graph Neural Networks: A survey

A General Benchmark Framework is Dynamic Graph Neural Network Need

A Comprehensive Survey on Graph Neural Networks

Dynamic Neural Networks: A Survey

Instant Graph Neural Networks for Dynamic Graphs

Performance Analysis of Graph Neural Network Frameworks

Graph Neural Networks for temporal graphs: State of the art, open challenges, and opportunities

Bridging the Gap between Spatial and Spectral Domains: A Survey on Graph Neural Networks

Adaptive Sampling Temporal Graph Network

Deep learning for dynamic graphs: models and benchmarks

Temporal Graph Networks for Deep Learning on Dynamic Graphs

Dynamic Spiking Framework for Graph Neural Networks

The Expressive Power of Graph Neural Networks: A Survey

A Survey of Graph Neural Networks in Real world: Imbalance, Noise, Privacy and OOD Challenges

Analysis of different temporal graph neural network configurations on dynamic graphs

From Continuous Dynamics to Graph Neural Networks: Neural Diffusion and Beyond

Bridging the Gap between Spatial and Spectral Domains: A Survey on Graph Neural Networks.