When Heterophily Meets Heterogeneous Graphs: Latent Graphs Guided Unsupervised Representation Learning

TL;DR

This paper introduces LatGRL, a novel framework for unsupervised heterogeneous graph representation learning that effectively handles semantic heterophily through latent graph construction and adaptive semantic fusion, validated by extensive experiments.

Contribution

The paper proposes a new framework, LatGRL, which addresses semantic heterophily in heterogeneous graphs using latent graphs and dual-frequency semantic fusion.

Findings

Effective in handling semantic heterophily in heterogeneous graphs

Scalable implementation for large-scale data

Validated on benchmark datasets with superior performance

Abstract

Unsupervised heterogeneous graph representation learning (UHGRL) has gained increasing attention due to its significance in handling practical graphs without labels. However, heterophily has been largely ignored, despite its ubiquitous presence in real-world heterogeneous graphs. In this paper, we define semantic heterophily and propose an innovative framework called Latent Graphs Guided Unsupervised Representation Learning (LatGRL) to handle this problem. First, we develop a similarity mining method that couples global structures and attributes, enabling the construction of fine-grained homophilic and heterophilic latent graphs to guide the representation learning. Moreover, we propose an adaptive dual-frequency semantic fusion mechanism to address the problem of node-level semantic heterophily. To cope with the massive scale of real-world data, we further design a scalable implementation. Extensive experiments on benchmark datasets validate the effectiveness and efficiency of our proposed framework. The source code and datasets have been made available at https://github.com/zxlearningdeep/LatGRL.