Heterogeneous Deep Graph Infomax

TL;DR

This paper introduces HDGI, an unsupervised method for learning high-quality node representations in heterogeneous graphs by maximizing local-global mutual information, improving performance on classification and clustering tasks.

Contribution

The paper proposes HDGI, a novel unsupervised GNN that leverages meta-paths, graph convolution, and attention to effectively learn heterogeneous graph representations.

Findings

HDGI outperforms state-of-the-art unsupervised methods in classification and clustering.

HDGI achieves comparable results to supervised GNNs in node classification.

Meta-paths and attention mechanisms enhance local and semantic representation learning.

Abstract

Graph representation learning is to learn universal node representations that preserve both node attributes and structural information. The derived node representations can be used to serve various downstream tasks, such as node classification and node clustering. When a graph is heterogeneous, the problem becomes more challenging than the homogeneous graph node learning problem. Inspired by the emerging information theoretic-based learning algorithm, in this paper we propose an unsupervised graph neural network Heterogeneous Deep Graph Infomax (HDGI) for heterogeneous graph representation learning. We use the meta-path structure to analyze the connections involving semantics in heterogeneous graphs and utilize graph convolution module and semantic-level attention mechanism to capture local representations. By maximizing local-global mutual information, HDGI effectively learns high-level node representations that can be utilized in downstream graph-related tasks. Experiment results show that HDGI remarkably outperforms state-of-the-art unsupervised graph representation learning methods on both classification and clustering tasks. By feeding the learned representations into a parametric model, such as logistic regression, we even achieve comparable performance in node classification tasks when comparing with state-of-the-art supervised end-to-end GNN models.