| Autor: |
Szyman, Paweł, Barbucha, Dariusz |
| Předmět: |
|
| Zdroj: |
Procedia Computer Science; 2024, Vol. 246, p4401-4410, 10p |
| Abstrakt: |
The paper focuses on the community detection problem in weighted social networks. The main goal of the paper is to experimentally verify whether assigning weights to links between vertices, indicating the strength of the relationship, enhances the performance of selected community detection algorithms compared to considering only the presence of the relation. A computational experiment has been carried out on the organizational social network of a public organization in Poland. It has been assumed that the structure of the network covers employees (represented by nodes) and relations reflecting digital communication between them (a link between two nodes exists if employees represented by these nodes exchanged at least a single email). Several algorithms have been tested using various performance measures for both unweighted and weighted social networks. The results of the study confirmed that considering weights on links between vertices leads to better performance of most tested algorithms and higher quality of detected communities. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Supplemental Index |
| Externí odkaz: |
|
