Complex Network Community Detection Based On Label Propagation And Genetic Algorithm

With the rapid development of Internet and social needs,the combination of various industries and the Internet has produced a large number of complex systems.Researchers abstract these complex systems into complex networks for research,revealing the laws and behaviors they contain,and exploring deep values and meanings.A common feature of complex networks is the community structure,which divides the network into clusters with tight internal connections and loose external connections.Community detection is a hot issue in recent years.Studying the community structure of complex networks helps us to analyze the common characteristics of individuals in the network and the relationship between the whole and the parts.This paper studies community detection from two aspects: single-objective optimization and multi-objective optimization.The main work is as follows:(1)This paper systematically introduces the research basis and progress of complex network community detection,classifies the commonly used complex network community detection algorithms,and analyzes the advantages and disadvantages of different community detection algorithms.It is found that most of these algorithms are not effective on large networks,and some algorithms must set a priori knowledge,which limits the usability of the algorithm.The genetic algorithm is described in detail,and the feasibility and applicability of genetic algorithm in community detection are analyzed.(2)A community detection algorithm based on modularity optimization is proposed.The traditional crossover operator and mutation operator are improved.The idea of label propagation is introduced into the crossover operator,and the generated community structure is used to cross and strengthen the algorithm.Local search capabilities.Control variation between the generated community structures and avoid invalid mutations.(3)In the framework of the standard NSGA2 algorithm,combined with the idea of label propagation,a multi-objective NSGA2 algorithm based on label propagation is proposed.The algorithm uses KKM and RC as the objective function,uses the characterbased coding method,randomly initializes the population,and uses the improved crossover operator and mutation operator to operate,thus realizing the evolution of the population.The simulation experiments are carried out in computer generated network and real network environment.The results show that compared with other community detection algorithms,the algorithm has the advantages of high resolution and strong search ability,and can effectively detect the community structure existing in complex networks.