Poria Pirozmand - Department of Computer Engineering, Aryan Institute of Science and Technology, Babol, Iran
MohammadReza Fadavi Amiri - Department of Computer Engineering, Shomal University, Amol ,Iran
Parisa Pirouzmand - Computer Science Department, Dalian University of Technology, Dalian, China

Overlapping Communities detection has become a critical step to understand the structure and dynamics of social networks in various area. However, traditional methods like Clique percolation ,local expansion and optimization methods, relatively new proposed link clustering and agent-based dynamic algorithms have inherent drawbacks in overlapping communities detection.Clique percolation only effective with high intensity networks. Node clustering is not adequate to capture pervasive overlaps, local expansion and optimization method seems more suitable to work with linear problem, while link clustering is criticized because of the high computational cost and unclear definition of communities. So, overlapping community detection is still a big challenge. In this paper, a Intelligent Swarm based Community Detection Method (SICDM) was proposed to discover the community overlapping. With the use of agents to determine the labels in the input network by taking into account the precision of different nodes in their decisions and utilization of knots as a way to analyze each Community with both a general view and tag-based independent determining forCommunity, the SICDM algorithms outperforms other similar algorithms in this area in terms of accuracy and performance. We also reviewed our algorithm in real-world networks where the structure of the real Community.The results suggested that SICDM was also capable of discovering the community structures in real networks with good behavior regarding of convergence.

Overlapping, detection, Intelligent Swarm, algorithms, Community