Graph neural network inspired algorithm for unsupervised
network community detection

J 2022

Graph neural network inspired algorithm for unsupervised network community detection

SOBOLEVSKY, Stanislav and Aliaksandr BELY

Basic information

Original name

Graph neural network inspired algorithm for unsupervised network community detection

Authors

SOBOLEVSKY, Stanislav (112 Belarus, guarantor, belonging to the institution) and Aliaksandr BELY (112 Belarus, belonging to the institution)

Edition

Applied Network Science, Springer Nature, 2022, 2364-8228

Other information

Language

English

Type of outcome

Článek v odborném periodiku

Field of Study

10200 1.2 Computer and information sciences

Country of publisher

United Kingdom of Great Britain and Northern Ireland

Confidentiality degree

není předmětem státního či obchodního tajemství

References:

URL

Impact factor

Impact factor: 2.200

RIV identification code

RIV/00216224:14310/22:00127675

Organization unit

Faculty of Science

DOI

http://dx.doi.org/10.1007/s41109-022-00500-z

UT WoS

000850086300002

Keywords in English

Complex networks; Community detection; Network science

Abstract

V originále

Network community detection often relies on optimizing partition quality functions, like modularity. This optimization appears to be a complex problem traditionally relying on discrete heuristics. And although the problem could be reformulated as continuous optimization, direct application of the standard optimization methods has limited efficiency in overcoming the numerous local extrema. However, the rise of deep learning and its applications to graphs offers new opportunities. And while graph neural networks have been used for supervised and unsupervised learning on networks, their application to modularity optimization has not been explored yet. This paper proposes a new variant of the recurrent graph neural network algorithm for unsupervised network community detection through modularity optimization. The new algorithm’s performance is compared against the state-of-the-art methods. The approach also serves as a proof-of-concept for the broader application of recurrent graph neural networks to unsupervised network optimization.

Links

EF16_019/0000822, research and development project

Name: Centrum excelence pro kyberkriminalitu, kyberbezpečnost a ochranu kritických informačních infrastruktur

MUNI/J/0008/2021, interní kód MU

Name: Digital City

Investor: Masaryk University, MASH JUNIOR - MUNI Award In Science and Humanities JUNIOR

Citovat

SOBOLEVSKY, Stanislav and Aliaksandr BELY. Graph neural network inspired algorithm for unsupervised network community detection. Applied Network Science. Springer Nature, 2022, vol. 7, No 1, p. 1-19. ISSN 2364-8228. Available from: https://dx.doi.org/10.1007/s41109-022-00500-z.

@article{2243485,
   author = {Sobolevsky, Stanislav and Bely, Aliaksandr},
   article_number = {1},
   doi = {http://dx.doi.org/10.1007/s41109-022-00500-z},
   keywords = {Complex networks; Community detection; Network science},
   language = {eng},
   issn = {2364-8228},
   journal = {Applied Network Science},
   title = {Graph neural network inspired algorithm for unsupervised network community detection},
   url = {https://appliednetsci.springeropen.com/articles/10.1007/s41109-022-00500-z},
   volume = {7},
   year = {2022}
}

TY  - JOUR
ID  - 2243485
AU  - Sobolevsky, Stanislav - Bely, Aliaksandr
PY  - 2022
TI  - Graph neural network inspired algorithm for unsupervised network community detection
JF  - Applied Network Science
VL  - 7
IS  - 1
SP  - 1-19
EP  - 1-19
PB  - Springer Nature
SN  - 23648228
KW  - Complex networks
KW  - Community detection
KW  - Network science
UR  - https://appliednetsci.springeropen.com/articles/10.1007/s41109-022-00500-z
N2  - Network community detection often relies on optimizing partition quality functions, like modularity. This optimization appears to be a complex problem traditionally relying on discrete heuristics. And although the problem could be reformulated as continuous optimization, direct application of the standard optimization methods has limited efficiency in overcoming the numerous local extrema. However, the rise of deep learning and its applications to graphs offers new opportunities. And while graph neural networks have been used for supervised and unsupervised learning on networks, their application to modularity optimization has not been explored yet. This paper proposes a new variant of the recurrent graph neural network algorithm for unsupervised network community detection through modularity optimization. The new algorithm’s performance is compared against the state-of-the-art methods. The approach also serves as a proof-of-concept for the broader application of recurrent graph neural networks to unsupervised network optimization.
ER  -

SOBOLEVSKY, Stanislav and Aliaksandr BELY. Graph neural network inspired algorithm for unsupervised network community detection. \textit{Applied Network Science}. Springer Nature, 2022, vol.~7, No~1, p.~1-19. ISSN~2364-8228. Available from: https://dx.doi.org/10.1007/s41109-022-00500-z.

Detailed Information on Publication Record