Novel objective-based coot puzzle optimisation for overlapping community expansion in complex networks
C.K. Suja, C.V.R. Harinarayanan, A. Arivalagan · International Journal of Networking and Virtual Organisations · 2024
Overlapping community detection receives more and more interest due to the possibility that certain nodes in real networks belong to numerous communities. However, the majority of the current overlapping community detection approaches significantly minimise the overfitting problems. This work proposes a novel approach for identifying influent nodes and overlapping community expansion in complex networks. Primarily, the identification of influential nodes is identified using the hybrid Renyi entropy and Mutual information (MI). Thereafter, in the overlapping community formation phase, the nodes other than the influential nodes are grouped to form an overlapping community. Finally, overlapping community expansion is accomplished utilising the proposed Coot Puzzle Optimisation Algorithm (CPOA). The proposed CPOA is the incorporation of the Puzzle Optimisation Algorithm (POA) with the Coot algorithm. The proposed CPOA has attained a value of 1.342, 0.549, and 0.634 for minimum Renyi-entropy-based betweenness, fitness of.