A Community detection technique for research collaboration networks based on frequent collaborators cores(Conference Paper)
- University of Novi Sad, Faculty of Sciences, Department of Mathematics and Informatics, Trg Dositeja Obradovića 4, Novi Sad, 21000, Serbia
Abstract
Community structure is one of prominent features of complex real-world networks. In this paper we propose a novel technique for detecting communities in research collaboration networks. The main idea of the algorithm is that research communities can be efficiently recovered from subgraphs encompassing frequent collaborators. Moreover, the algorithm can be used to cluster weighted undirected networks from other domains as well. An experimental evaluation of the algorithm was conducted on a co-authorship network representing collaborations between researchers employed at our Department. The results of the evaluation showed that the algorithm identifies strong and meaningful clusters corresponding to groups dealing with specific research topics. Moreover, we compared our method to seven other community detection techniques showing that it performs better or equally with respect to the quality of obtained community structures. © 2016 ACM.
Indexed keywords
| Engineering controlled terms: | Population dynamics |
|---|---|
| Engineering uncontrolled terms | Co-authorship networksCommunity detectionCommunity structuresExperimental evaluationFrequent collaboratorsReal-world networksResearch collaborationsResearch communities |
| Engineering main heading: | Clustering algorithms |
- ISBN: 978-145033739-7
- Source Type: Conference Proceeding
- Original language: English
- DOI: 10.1145/2851613.2851809
- Document Type: Conference Paper
- Sponsors: ACM Special Interest Group on Applied Computing (SIGAPP)
- Publisher: Association for Computing Machinery
© Copyright 2019 Elsevier B.V., All rights reserved.
Cited by 3 documents
(2022) IEEE Transactions on Computational Social Systems
(2019) Computing
(2019) Intelligent Systems Reference Library