Benchmarking Graph Neural Networks on Dynamic Link Prediction

Joakim Skarding, Matthew Hellmich, Bogdan Gabrys, Katarzyna Musial-Gabrys

Keywords: Graph neural network, dynamic graph neural network, link prediction, dynamic link prediction, temporal graph

Abstract: Graph neural networks (GNNs) are rapidly becoming the dominant way to learn on graph-structured data. Link prediction is a near-universal benchmark for new GNN models. Many advanced models such as Dynamic graph neural networks(DGNNs) specifically target dynamic link prediction. However, these models, particularly DGNNs, are rarely compared to each other or existing heuristics. Different works evaluate their models in different ways, thus one cannot compare evaluation metrics directly. Motivated by this, we perform a comprehensive comparison study. We compare link prediction heuristics, GNNs, discrete DGNNs, and continuous DGNNs on dynamic link prediction. We find that simple link prediction heuristics often perform better than GNNs and DGNNs, different sliding window sizes greatly affect performance, and of all examined graph neural networks, that DGNNs consistently outperform static GNNs.

One-sentence Summary: A fair comparison of dynamic graph neural networks on the dynamic link prediction task.