No Cover Image

Journal article 210 views

Graph Contrastive Learning for Tracking Dynamic Communities in Temporal Networks

Yun Ai, Xianghua Xie Orcid Logo, Xiaoke Ma Orcid Logo

IEEE Transactions on Emerging Topics in Computational Intelligence, Volume: 8, Issue: 5, Pages: 3422 - 3435

Swansea University Author: Xianghua Xie Orcid Logo

Full text not available from this repository: check for access using links below.

Check full text

DOI (Published version): 10.1109/tetci.2024.3386844

Abstract

Temporal networks are ubiquitous because complex systems in nature and society are evolving, and tracking dynamic communities is critical for revealing the mechanism of systems. Moreover, current algorithms utilize temporal smoothness framework to balance clustering accuracy at current time and clus...

Full description

Published in: IEEE Transactions on Emerging Topics in Computational Intelligence
ISSN: 2471-285X
Published: Institute of Electrical and Electronics Engineers (IEEE) 2024
Online Access: Check full text

URI: https://cronfa.swan.ac.uk/Record/cronfa69415
first_indexed 2025-05-02T14:52:43Z
last_indexed 2025-06-20T04:58:14Z
id cronfa69415
recordtype SURis
fullrecord <?xml version="1.0"?><rfc1807><datestamp>2025-06-19T10:18:38.9262228</datestamp><bib-version>v2</bib-version><id>69415</id><entry>2025-05-02</entry><title>Graph Contrastive Learning for Tracking Dynamic Communities in Temporal Networks</title><swanseaauthors><author><sid>b334d40963c7a2f435f06d2c26c74e11</sid><ORCID>0000-0002-2701-8660</ORCID><firstname>Xianghua</firstname><surname>Xie</surname><name>Xianghua Xie</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2025-05-02</date><deptcode>MACS</deptcode><abstract>Temporal networks are ubiquitous because complex systems in nature and society are evolving, and tracking dynamic communities is critical for revealing the mechanism of systems. Moreover, current algorithms utilize temporal smoothness framework to balance clustering accuracy at current time and clustering drift at historical time, which are criticized for failing to characterize the temporality of networks and determine its importance. To overcome these problems, we propose a novel algorithm by joining Non-negative matrix factorization and Contrastive learning for Dynamic Community detection (jNCDC). Specifically, jNCDC learns the features of vertices by projecting successive snapshots into a shared subspace to learn the low-dimensional representation of vertices with matrix factorization. Subsequently, it constructs an evolution graph to explicitly measure relations of vertices by representing vertices at current time with features at historical time, paving a way to characterize the dynamics of networks at the vertex-level. Finally, graph contrastive learning utilizes the roles of vertices to select positive and negative samples to further improve the quality of features. These procedures are seamlessly integrated into an overall objective function, and optimization rules are deduced. To the best of our knowledge, jNCDC is the first graph contrastive learning for dynamic community detection, that provides an alternative for the current temporal smoothness framework. Experimental results demonstrate that jNCDC is superior to the state-of-the-art approaches in terms of accuracy.</abstract><type>Journal Article</type><journal>IEEE Transactions on Emerging Topics in Computational Intelligence</journal><volume>8</volume><journalNumber>5</journalNumber><paginationStart>3422</paginationStart><paginationEnd>3435</paginationEnd><publisher>Institute of Electrical and Electronics Engineers (IEEE)</publisher><placeOfPublication/><isbnPrint/><isbnElectronic/><issnPrint/><issnElectronic>2471-285X</issnElectronic><keywords/><publishedDay>17</publishedDay><publishedMonth>4</publishedMonth><publishedYear>2024</publishedYear><publishedDate>2024-04-17</publishedDate><doi>10.1109/tetci.2024.3386844</doi><url/><notes/><college>COLLEGE NANME</college><department>Mathematics and Computer Science School</department><CollegeCode>COLLEGE CODE</CollegeCode><DepartmentCode>MACS</DepartmentCode><institution>Swansea University</institution><apcterm/><funders>This work was supported in part by the National Natural Science Foundation of China under Grant 62272361 and in part by the Shaanxi Natural Science Funds for Distinguished Young Scholar Program under Grant 2022JC-38.</funders><projectreference/><lastEdited>2025-06-19T10:18:38.9262228</lastEdited><Created>2025-05-02T15:49:37.6998437</Created><path><level id="1">Faculty of Science and Engineering</level><level id="2">School of Mathematics and Computer Science - Computer Science</level></path><authors><author><firstname>Yun</firstname><surname>Ai</surname><order>1</order></author><author><firstname>Xianghua</firstname><surname>Xie</surname><orcid>0000-0002-2701-8660</orcid><order>2</order></author><author><firstname>Xiaoke</firstname><surname>Ma</surname><orcid>0000-0002-5604-7137</orcid><order>3</order></author></authors><documents/><OutputDurs/></rfc1807>
spelling 2025-06-19T10:18:38.9262228 v2 69415 2025-05-02 Graph Contrastive Learning for Tracking Dynamic Communities in Temporal Networks b334d40963c7a2f435f06d2c26c74e11 0000-0002-2701-8660 Xianghua Xie Xianghua Xie true false 2025-05-02 MACS Temporal networks are ubiquitous because complex systems in nature and society are evolving, and tracking dynamic communities is critical for revealing the mechanism of systems. Moreover, current algorithms utilize temporal smoothness framework to balance clustering accuracy at current time and clustering drift at historical time, which are criticized for failing to characterize the temporality of networks and determine its importance. To overcome these problems, we propose a novel algorithm by joining Non-negative matrix factorization and Contrastive learning for Dynamic Community detection (jNCDC). Specifically, jNCDC learns the features of vertices by projecting successive snapshots into a shared subspace to learn the low-dimensional representation of vertices with matrix factorization. Subsequently, it constructs an evolution graph to explicitly measure relations of vertices by representing vertices at current time with features at historical time, paving a way to characterize the dynamics of networks at the vertex-level. Finally, graph contrastive learning utilizes the roles of vertices to select positive and negative samples to further improve the quality of features. These procedures are seamlessly integrated into an overall objective function, and optimization rules are deduced. To the best of our knowledge, jNCDC is the first graph contrastive learning for dynamic community detection, that provides an alternative for the current temporal smoothness framework. Experimental results demonstrate that jNCDC is superior to the state-of-the-art approaches in terms of accuracy. Journal Article IEEE Transactions on Emerging Topics in Computational Intelligence 8 5 3422 3435 Institute of Electrical and Electronics Engineers (IEEE) 2471-285X 17 4 2024 2024-04-17 10.1109/tetci.2024.3386844 COLLEGE NANME Mathematics and Computer Science School COLLEGE CODE MACS Swansea University This work was supported in part by the National Natural Science Foundation of China under Grant 62272361 and in part by the Shaanxi Natural Science Funds for Distinguished Young Scholar Program under Grant 2022JC-38. 2025-06-19T10:18:38.9262228 2025-05-02T15:49:37.6998437 Faculty of Science and Engineering School of Mathematics and Computer Science - Computer Science Yun Ai 1 Xianghua Xie 0000-0002-2701-8660 2 Xiaoke Ma 0000-0002-5604-7137 3
title Graph Contrastive Learning for Tracking Dynamic Communities in Temporal Networks
spellingShingle Graph Contrastive Learning for Tracking Dynamic Communities in Temporal Networks
Xianghua Xie
title_short Graph Contrastive Learning for Tracking Dynamic Communities in Temporal Networks
title_full Graph Contrastive Learning for Tracking Dynamic Communities in Temporal Networks
title_fullStr Graph Contrastive Learning for Tracking Dynamic Communities in Temporal Networks
title_full_unstemmed Graph Contrastive Learning for Tracking Dynamic Communities in Temporal Networks
title_sort Graph Contrastive Learning for Tracking Dynamic Communities in Temporal Networks
author_id_str_mv b334d40963c7a2f435f06d2c26c74e11
author_id_fullname_str_mv b334d40963c7a2f435f06d2c26c74e11_***_Xianghua Xie
author Xianghua Xie
author2 Yun Ai
Xianghua Xie
Xiaoke Ma
format Journal article
container_title IEEE Transactions on Emerging Topics in Computational Intelligence
container_volume 8
container_issue 5
container_start_page 3422
publishDate 2024
institution Swansea University
issn 2471-285X
doi_str_mv 10.1109/tetci.2024.3386844
publisher Institute of Electrical and Electronics Engineers (IEEE)
college_str Faculty of Science and Engineering
hierarchytype
hierarchy_top_id facultyofscienceandengineering
hierarchy_top_title Faculty of Science and Engineering
hierarchy_parent_id facultyofscienceandengineering
hierarchy_parent_title Faculty of Science and Engineering
department_str School of Mathematics and Computer Science - Computer Science{{{_:::_}}}Faculty of Science and Engineering{{{_:::_}}}School of Mathematics and Computer Science - Computer Science
document_store_str 0
active_str 0
description Temporal networks are ubiquitous because complex systems in nature and society are evolving, and tracking dynamic communities is critical for revealing the mechanism of systems. Moreover, current algorithms utilize temporal smoothness framework to balance clustering accuracy at current time and clustering drift at historical time, which are criticized for failing to characterize the temporality of networks and determine its importance. To overcome these problems, we propose a novel algorithm by joining Non-negative matrix factorization and Contrastive learning for Dynamic Community detection (jNCDC). Specifically, jNCDC learns the features of vertices by projecting successive snapshots into a shared subspace to learn the low-dimensional representation of vertices with matrix factorization. Subsequently, it constructs an evolution graph to explicitly measure relations of vertices by representing vertices at current time with features at historical time, paving a way to characterize the dynamics of networks at the vertex-level. Finally, graph contrastive learning utilizes the roles of vertices to select positive and negative samples to further improve the quality of features. These procedures are seamlessly integrated into an overall objective function, and optimization rules are deduced. To the best of our knowledge, jNCDC is the first graph contrastive learning for dynamic community detection, that provides an alternative for the current temporal smoothness framework. Experimental results demonstrate that jNCDC is superior to the state-of-the-art approaches in terms of accuracy.
published_date 2024-04-17T05:23:52Z
_version_ 1851641176165187584
score 11.089822

Similar Items