Input Snapshots Fusion for Scalable Discrete-Time Dynamic Graph Neural Networks

TL;DR

This paper introduces SFDyG, a novel dynamic graph neural network that fuses multiple snapshots into one graph, reducing computational costs and improving future link prediction accuracy on large-scale discrete-time dynamic graphs.

Contribution

The paper proposes a new method combining Hawkes processes with GNNs to efficiently model temporal and structural patterns by fusing snapshots, decoupling complexity from snapshot count.

Findings

Outperforms existing methods on eight datasets.

Efficient training for large-scale dynamic graphs.

Improves future link prediction accuracy.

Abstract

In recent years, there has been a surge in research on dynamic graph representation learning, primarily focusing on modeling the evolution of temporal-spatial patterns in real-world applications. However, within the domain of discrete-time dynamic graphs, the exploration of temporal edges remains underexplored. Existing approaches often rely on additional sequential models to capture dynamics, leading to high computational and memory costs, particularly for large-scale graphs. To address this limitation, we propose the Input {\bf S}napshots {\bf F}usion based {\bf Dy}namic {\bf G}raph Neural Network (SFDyG), which combines Hawkes processes with graph neural networks to capture temporal and structural patterns in dynamic graphs effectively. By fusing multiple snapshots into a single temporal graph, SFDyG decouples computational complexity from the number of snapshots, enabling efficient full-batch and mini-batch training. Experimental evaluations on eight diverse dynamic graph datasets for future link prediction tasks demonstrate that SFDyG consistently outperforms existing methods.