Deep graph representation learning: methods, applications, and challenges

Graph representation learning has emerged as a crucial research area in recent years, aiming to generate vector representations that accurately capture the structure and features of graphs. These vectors play a vital role in downstream tasks such as node classification, link prediction, and anomaly...

Full description

Saved in:
Bibliographic Details
Main Authors: ZHANG Xulong, QU Xiaoyang, XIAO Chunguang, WANG Jianzong
Format: Article
Language:zho
Published: China InfoCom Media Group 2025-01-01
Series:大数据
Subjects:
Online Access:http://www.j-bigdataresearch.com.cn/thesisDetails?columnId=109257984&Fpath=home&index=0
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850136777252143104
author ZHANG Xulong
QU Xiaoyang
XIAO Chunguang
WANG Jianzong
author_facet ZHANG Xulong
QU Xiaoyang
XIAO Chunguang
WANG Jianzong
author_sort ZHANG Xulong
collection DOAJ
description Graph representation learning has emerged as a crucial research area in recent years, aiming to generate vector representations that accurately capture the structure and features of graphs. These vectors play a vital role in downstream tasks such as node classification, link prediction, and anomaly detection. This paper presents a comprehensive survey of graph representation learning methods, categorizing them into traditional graph embedding methods and Graph Neural Network (GNN) based approaches. We discuss various techniques within these categories, including matrix factorization, random walks, graph convolutional networks, and graph Transformers. Furthermore, we delve into the specific applications of GNN in heterogeneous graph embedding, encompassing both static and dynamic aspects. This paper also explores the challenges and future directions of graph representation learning, including scalability and dynamics.
format Article
id doaj-art-fe99ae0f84f14a4c8bff354524474c3e
institution OA Journals
issn 2096-0271
language zho
publishDate 2025-01-01
publisher China InfoCom Media Group
record_format Article
series 大数据
spelling doaj-art-fe99ae0f84f14a4c8bff354524474c3e2025-08-20T02:31:02ZzhoChina InfoCom Media Group大数据2096-02712025-01-01123109257984Deep graph representation learning: methods, applications, and challengesZHANG XulongQU XiaoyangXIAO ChunguangWANG JianzongGraph representation learning has emerged as a crucial research area in recent years, aiming to generate vector representations that accurately capture the structure and features of graphs. These vectors play a vital role in downstream tasks such as node classification, link prediction, and anomaly detection. This paper presents a comprehensive survey of graph representation learning methods, categorizing them into traditional graph embedding methods and Graph Neural Network (GNN) based approaches. We discuss various techniques within these categories, including matrix factorization, random walks, graph convolutional networks, and graph Transformers. Furthermore, we delve into the specific applications of GNN in heterogeneous graph embedding, encompassing both static and dynamic aspects. This paper also explores the challenges and future directions of graph representation learning, including scalability and dynamics.http://www.j-bigdataresearch.com.cn/thesisDetails?columnId=109257984&Fpath=home&index=0graph neural networksgraph embeddingheterogeneous graphdeep learning
spellingShingle ZHANG Xulong
QU Xiaoyang
XIAO Chunguang
WANG Jianzong
Deep graph representation learning: methods, applications, and challenges
大数据
graph neural networks
graph embedding
heterogeneous graph
deep learning
title Deep graph representation learning: methods, applications, and challenges
title_full Deep graph representation learning: methods, applications, and challenges
title_fullStr Deep graph representation learning: methods, applications, and challenges
title_full_unstemmed Deep graph representation learning: methods, applications, and challenges
title_short Deep graph representation learning: methods, applications, and challenges
title_sort deep graph representation learning methods applications and challenges
topic graph neural networks
graph embedding
heterogeneous graph
deep learning
url http://www.j-bigdataresearch.com.cn/thesisDetails?columnId=109257984&Fpath=home&index=0
work_keys_str_mv AT zhangxulong deepgraphrepresentationlearningmethodsapplicationsandchallenges
AT quxiaoyang deepgraphrepresentationlearningmethodsapplicationsandchallenges
AT xiaochunguang deepgraphrepresentationlearningmethodsapplicationsandchallenges
AT wangjianzong deepgraphrepresentationlearningmethodsapplicationsandchallenges