Dual Hierarchical Attention Networks for Bi-typed Heterogeneous Graph Learning

by   Yu Zhao, et al.
Beihang University
Southwestern University of Finance and Economics

Bi-typed heterogeneous graphs are applied in many real-world scenarios. However, previous heterogeneous graph learning studies usually ignore the complex interactions among the bi-typed entities in such heterogeneous graphs. To address this issue, in this paper we propose a novel Dual Hierarchical Attention Networks (DHAN) to learn comprehensive node representations on the bi-typed heterogeneous graphs with intra-class and inter-class hierarchical attention networks. Specifically, the intra-class attention aims to learn the node representations from its same type of neighbors, while inter-class attention is able to aggregate node representations from its different types of neighbors. Hence, the dual attention operations enable DHAN to sufficiently leverage not only the node intra-class neighboring information but also the inter-class neighboring information in the bi-typed heterogeneous graph. Experimental results on various tasks against the state-of-the-arts sufficiently confirm the capability of DHAN in learning node comprehensive representations on the bi-typed heterogeneous


page 2

page 3

page 5

page 6

page 7

page 8

page 9

page 10


Heterogeneous Temporal Graph Neural Network

Graph neural networks (GNNs) have been broadly studied on dynamic graphs...

Not All Neighbors Are Worth Attending to: Graph Selective Attention Networks for Semi-supervised Learning

Graph attention networks (GATs) are powerful tools for analyzing graph d...

Contextual Heterogeneous Graph Network for Human-Object Interaction Detection

Human-object interaction(HOI) detection is an important task for underst...

HAGNN: Hybrid Aggregation for Heterogeneous Graph Neural Networks

Heterogeneous graph neural networks (GNNs) have been successful in handl...

Hybrid Micro/Macro Level Convolution for Heterogeneous Graph Learning

Heterogeneous graphs are pervasive in practical scenarios, where each gr...

GripNet: Graph Information Propagation on Supergraph for Heterogeneous Graphs

Heterogeneous graph representation learning aims to learn low-dimensiona...

HetNetAligner: Design and Implementation of an algorithm for heterogeneous network alignment on Apache Spark

The importance of the use of networks to model and analyse biological da...

Please sign up or login with your details

Forgot password? Click here to reset