| A. Kanavos, M. Trigka, E. Dritsas, G. Vonitsanos, Ph. Mylonas |
| Community Detection Algorithms for Cultural and Natural Heritage Data in Social Networks |
| 10th Mining Humanistic Data Workshop (MHDW 2021) in conjunction with 17th AIAI Conference, Crete, Greece, June 2021 |
|
ABSTRACT
|
| In recent years, researchers have employed various approaches and techniques to identify community structures in social networks. Community discovery is an essential topic in social network analysis since it provides a way for recursively decomposing a large social graph to easily interpretable subgraphs. The implementation of four major community discovery algorithms, namely the Breadth-First Search, the Louvain, the MaxToMin, and the Propinquity Dynamics, is described. Their correctness was functionally evaluated in the four most widely used graphs with vastly different characteristics and a dataset retrieved from Twitter regarding cultural heritage data. The primary finding was that the Propinquity Dynamics algorithm outperforms the other algorithms in terms of NMI for most graphs. In contrast, this algorithm with the Louvain performs almost the same regarding modularity.
|
| 25 June , 2021 |
| A. Kanavos, M. Trigka, E. Dritsas, G. Vonitsanos, Ph. Mylonas, "Community Detection Algorithms for Cultural and Natural Heritage Data in Social Networks", 10th Mining Humanistic Data Workshop (MHDW 2021) in conjunction with 17th AIAI Conference, Crete, Greece, June 2021 |
[  PDF] [
BibTex] [
Print] [
 Back] |
