A Discrete Bat Algorithm for Detecting Overlapping Communities in Dynamic Graph-Based Social Networks

Community identity is an important task in analyzing social networks, which aims to identify underghier, where nodes are closely connected to internal and are very connected to the rest of the network. These communities can be dissatisfied - where each node belongs to only one society - or overlapping, leaves the nodes related to many communities. The study suggests an unspoilt BAT algorithm (DBA) to detect local communities in symmetrical dynamic networks. Inspired by the eco -location behavior of microbhates, the original BAT algorithm effectively solves continuous adaptation problems, but the application is limited in disconnected domains such as societal identity. To address this, a discreet adaptation has been developed, which represents the status of the bat in an appropriate way to solve combinatory problems. Social identification is designed as a multi -use problem in the stages of a dynamic network time, with two objective functions: First, stable and meaningful social structures mean each time, and the other promotes temporary stability by maximizing the resemblance in continuous time stages. Experimental results suggest that the proposed DBA has rejected existing methods in existing methods, including the Particle Herd-Adaptation (PSO), Genetic Algorithms (GA) and Multi-Lens Biogography-based optimization (MOBO), the lowest bordered rand (ARI) and the lowest error in 98.