Saved in:
Bibliographic Details
Main Authors: Li, Chao, Guo, Zijie, He, Qiuting, Xu, Hao, He, Kun
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2307.08430
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916661050736640
author Li, Chao
Guo, Zijie
He, Qiuting
Xu, Hao
He, Kun
author_facet Li, Chao
Guo, Zijie
He, Qiuting
Xu, Hao
He, Kun
contents Utilizing long-range dependency, a concept extensively studied in homogeneous graphs, remains underexplored in heterogeneous graphs, especially on large ones, posing two significant challenges: Reducing computational costs while maximizing effective information utilization in the presence of heterogeneity, and overcoming the over-smoothing issue in graph neural networks. To address this gap, we investigate the importance of different meta-paths and introduce an automatic framework for utilizing long-range dependency on heterogeneous graphs, denoted as Long-range Meta-path Search through Progressive Sampling (LMSPS). Specifically, we develop a search space with all meta-paths related to the target node type. By employing a progressive sampling algorithm, LMSPS dynamically shrinks the search space with hop-independent time complexity. Through a sampling evaluation strategy, LMSPS conducts a specialized and effective meta-path selection, leading to retraining with only effective meta-paths, thus mitigating costs and over-smoothing. Extensive experiments across diverse heterogeneous datasets validate LMSPS's capability in discovering effective long-range meta-paths, surpassing state-of-the-art methods. Our code is available at https://github.com/JHL-HUST/LMSPS.
format Preprint
id arxiv_https___arxiv_org_abs_2307_08430
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Long-range Meta-path Search on Large-scale Heterogeneous Graphs
Li, Chao
Guo, Zijie
He, Qiuting
Xu, Hao
He, Kun
Artificial Intelligence
Utilizing long-range dependency, a concept extensively studied in homogeneous graphs, remains underexplored in heterogeneous graphs, especially on large ones, posing two significant challenges: Reducing computational costs while maximizing effective information utilization in the presence of heterogeneity, and overcoming the over-smoothing issue in graph neural networks. To address this gap, we investigate the importance of different meta-paths and introduce an automatic framework for utilizing long-range dependency on heterogeneous graphs, denoted as Long-range Meta-path Search through Progressive Sampling (LMSPS). Specifically, we develop a search space with all meta-paths related to the target node type. By employing a progressive sampling algorithm, LMSPS dynamically shrinks the search space with hop-independent time complexity. Through a sampling evaluation strategy, LMSPS conducts a specialized and effective meta-path selection, leading to retraining with only effective meta-paths, thus mitigating costs and over-smoothing. Extensive experiments across diverse heterogeneous datasets validate LMSPS's capability in discovering effective long-range meta-paths, surpassing state-of-the-art methods. Our code is available at https://github.com/JHL-HUST/LMSPS.
title Long-range Meta-path Search on Large-scale Heterogeneous Graphs
topic Artificial Intelligence
url https://arxiv.org/abs/2307.08430