| Title: | How should we benchmark community detection algorithms in complex networks? |
|---|
| Authors: | ID Pritržnik, Robi (Author) |
|---|
| Files: |  RAZ_Pritrznik_Robi_2025.pdf (12,52 MB)
MD5: 2EBEEB33F12E23AE224DC803CB606675
|
|---|
| Language: | English |
|---|
| Work type: | Unknown |
|---|
| Typology: | 1.08 - Published Scientific Conference Contribution |
|---|
| Organization: | FIŠ - Faculty of Information Studies in Novo mesto
|
|---|
| Abstract: | In this article we discuss how should we benchmark community detection algorithms in complex networks. We compare the community detection algorithms Louvain, Leiden, Label Propagation, Fast Label Propagation, Greedy modularity, Infomap, Walktrap and Girvan-Newman on complex networks of the Zachary karate club, Synthetic Network, a social network from X (Twitter), a neuroscience network, a email network and a patent citation network in the USA. We find that the speed of algorithms depends on the size and structure of networks. It turns out that among the considered algorithms for community detection in large networks, the Leiden algorithm is the most suitable, while on average the Fast Label Propagation algorithm performed the fastest in all cases. It is shown that on LFR benchmark network, algorithms successfully detect the same number of communities, however when we apply the same algorithms on complex networks the results are variable based on specific algorithm. |
|---|
| Keywords: | community detection, networks and graphs, network analysis, complex networks |
|---|
| Publication status: | Published |
|---|
| Publication version: | Version of Record |
|---|
| Publication date: | 16.01.2026 |
|---|
| Year of publishing: | 2025 |
|---|
| Number of pages: | Str. [88-97] |
|---|
| PID: | 20.500.12556/ReVIS-13041  |
|---|
| COBISS.SI-ID: | 264969219 |
|---|
| UDC: | 004.8:519.17 |
|---|
| Note: | Nasl. z nasl. zaslona;
Opis vira z dne 16. 1. 2026;
|
|---|
| Publication date in ReVIS: | 22.01.2026 |
|---|
| Views: | 41 |
|---|
| Downloads: | 0 |
|---|
| Metadata: |    |
|---|
|
:
|
Copy citation |
|---|
| | | | Share: |  |
|---|
Hover the mouse pointer over a document title to show the abstract or click
on the title to get all document metadata. |
