Neighborhood Overlap-Aware High-Order Graph Neural Network for Dynamic Graph Learning

TL;DR

This paper introduces NO-HGNN, a novel dynamic graph neural network that incorporates neighborhood overlap information into high-order message passing to better model complex structural patterns in evolving graphs.

Contribution

The paper proposes a neighborhood overlap-aware high-order GNN that explicitly models complex structural dependencies in dynamic graphs, improving link prediction performance.

Findings

NO-HGNN outperforms state-of-the-art methods in link prediction.

Neighborhood overlap significantly enhances modeling of node interactions.

The approach effectively captures complex structural patterns in dynamic graphs.

Abstract

Dynamic graph learning (DGL) aims to learn informative and temporally-evolving node embeddings to support downstream tasks such as link prediction. A fundamental challenge in DGL lies in effectively modeling both the temporal dynamics and structural dependencies of evolving graph topologies. Recent advances in Dynamic Graph Neural Networks (DGNNs) have obtained remarkable success by leveraging message-passing mechanisms to capture pairwise node interactions. However, these approaches often overlook more complex structural patterns, particularly neighborhood overlap, which can play a critical role in characterizing node interactions. To overcome this limitation, we introduce the Neighborhood Overlap-Aware High-Order Graph Neural Network (NO-HGNN), which is built upon two key innovations: (a) computing a correlation score based on the extent of neighborhood overlap to better capture complex node interactions; and (b) embedding this correlation directly into the message-passing process of high-order graph neural networks in the DGL. Experiments on two real-world dynamic graphs show that NO-HGNN achieves notable improvements in link prediction accuracy, outperforming several state-of-the-art approaches.