Name
Abstract | ||
|---|---|---|
Local community detection (LCD for short) aims at finding a community structure in a network starting from a seed (i.e., a “local” starting vertex). In a process of LCD, local community metrics are crucial since they serve as the measurements for the quality of the detected local community. Even if various algorithms have been proposed for LCD, there has been few investigation on the key features of these local community metrics, resulting in a lack of guidelines on how to choose these metrics in practice. To make up this inadequacy, this paper first investigates the effectiveness and efficiency of local community metrics via LCD accuracy comparison and scalability study, and then studies the insensitivity of these metrics to different seeds in a target community structure, followed by evaluating their performance on local communities with noisy vertices inside. In addition, a set of guidelines for the selection of local community metrics are given based on our findings concluded from extensive experiments. |
Year | DOI | Venue |
|---|---|---|
2018 | https://doi.org/10.1007/s10844-017-0480-5 | J. Intell. Inf. Syst. |
Keywords | Field | DocType |
Local community metrics,Accuracy,Efficiency,Seed-insensitivity,Noise impact | Local community,Data mining,Community structure,Computer science,Artificial intelligence,Machine learning,Scalability | Journal |
Volume | Issue | ISSN |
51 | 1 | 0925-9902 |
Citations | PageRank | References |
0 | 0.34 | 12 |
Authors | ||
4 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Lianhang Ma | 1 | 58 | 3.96 |
| Kevin Chiew | 2 | 116 | 11.06 |
| Hao Huang | 3 | 589 | 104.49 |
| Qinming He | 4 | 371 | 41.53 |