Multi-Hyperbolic Space-based Heterogeneous Graph Attention Network

TL;DR

This paper introduces MSGAT, a novel heterogeneous graph neural network that employs multiple hyperbolic spaces to better model the diverse power-law structures inherent in complex heterogeneous graphs, outperforming existing methods.

Contribution

The paper proposes a multi-hyperbolic space approach for heterogeneous graph embedding, addressing limitations of single hyperbolic space models in capturing diverse graph structures.

Findings

MSGAT outperforms state-of-the-art baselines in multiple tasks.

Effectively captures complex structures of heterogeneous graphs.

Demonstrates the advantage of using multiple hyperbolic spaces.

Abstract

To leverage the complex structures within heterogeneous graphs, recent studies on heterogeneous graph embedding use a hyperbolic space, characterized by a constant negative curvature and exponentially increasing space, which aligns with the structural properties of heterogeneous graphs. However, despite heterogeneous graphs inherently possessing diverse power-law structures, most hyperbolic heterogeneous graph embedding models use a single hyperbolic space for the entire heterogeneous graph, which may not effectively capture the diverse power-law structures within the heterogeneous graph. To address this limitation, we propose Multi-hyperbolic Space-based heterogeneous Graph Attention Network (MSGAT), which uses multiple hyperbolic spaces to effectively capture diverse power-law structures within heterogeneous graphs. We conduct comprehensive experiments to evaluate the effectiveness of MSGAT. The experimental results demonstrate that MSGAT outperforms state-of-the-art baselines in various graph machine learning tasks, effectively capturing the complex structures of heterogeneous graphs.