Parallel Algorithm for Identifying Overlapping Communities from Local Extension

DOI

10.2991/aiie-16.2016.22How to use a DOI?

Keywords

overlapping community; clustering coefficient; core node; parallel computing; overlap degree; modularty increment

Abstract

The overlapping communities identification in complex networks was studied with the parallel algorithm from local extension. The algorithm thought originated from clustering of modularity. Firstly, based on the Map/Reduce framework, the core node was selected according to the distribution and aggregation of nodes; Then, the algorithm calculated the members of newly added nodes and the existing small community formed by core node and neighbor nodes, and achieved a gradual expansion from local community to global community; Finally, the overlapping degree and modularity increment were used to effectively identify community. Experimental results show that: Community identification algorithm realized an extension from local community, and improved the effect of aggregation, and its parallel computing is faster and more efficient than traditional algorithms.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Hua Long
AU  - Baoan Li
PY  - 2016/11
DA  - 2016/11
TI  - Parallel Algorithm for Identifying Overlapping Communities from Local Extension
BT  - Proceedings of the 2016 2nd International Conference on Artificial Intelligence and Industrial Engineering (AIIE 2016)
PB  - Atlantis Press
SP  - 90
EP  - 93
SN  - 1951-6851
UR  - https://doi.org/10.2991/aiie-16.2016.22
DO  - 10.2991/aiie-16.2016.22
ID  - Long2016/11
ER  -