Representation Learning on Heterophilic Graph with Directional Neighborhood Attention

arXiv:2403.01475v1 Announce Type: new
Abstract: Graph Attention Network (GAT) is one of the most popular Graph Neural Network (GNN) architecture, which employs the attention mechanism to learn edge weights and has demonstrated promising performance in various applications. However, since it only incorporates information from immediate neighborhood, it lacks the ability to capture long-range and global graph information, leading to unsatisfactory performance on some datasets, particularly on heterophilic graphs. To address this limitation, we propose the Directional Graph Attention Network (DGAT) …

abstract applications architecture arxiv attention cs.ai cs.lg cs.si edge gnn graph graph neural network information learn network neural network performance popular representation representation learning type