Community Detection In Complex Networks Based On Parallel Clustering Analysis

With the fast development of Internet technology,big data has gradually become the focus and difficulty of computing field.It has very important theoretical significance and research value.Community detection in complex networks is a very important research topic.In order to achieve higher precision and more efficient community detection for complex networks,this paper mainly uses Hadoop parallel computing platform for algorithm parallelization.The purpose is to effectively solve the problem of community detection in large and complex networks.Firstly,the Hadoop platform is used to design the distributed parallel computing model of complex network community discovery process,and the parallel community detection division of complex network original data is divided by HDFS model.Then the parallel batch processing advantage of HDFS model is combined to realize the parallel processing of the complex network community discovery process.(1)For complex network community detection,the Hadoop platform is applied to realize the distributed community parallel detection of complex networks.(2)Based on the principle of Cluster-dp method,a center decentralized sparse selection method based on rank is designed to improve the initial seed selection process of community discovery process and achieve good results.Combined with the traditional K-means clustering algorithm,we classify the nodes in complex community network to achieve effective detection of community structure.Aiming at the sparse relationship of the individual in the network community,a clustering algorithm,called the EkNN algorithm,used to strengthen the individual attributes of the community nodes is designed,to reduce the impact of the data noise on the community discovery process(3)In order to improve the algorithm performance of the community discovery process,a complex network community detection method based on the parallel KNN algorithm is taken into consideration for the overlapping structure in complex network community.A parallel clustering algorithm based on parallel framework is designed to improve the efficiency of community discovery process.(4)Aiming at the complex network community detection problem,this paper improves the association rule algorithm by solving the computation redundancy problem in the association rule algorithm,and designs a community detection association rule algorithm with single constraint.In order to reduce the computational complexity of the complex network community detection algorithm,this paper combines the Mapreduce model in the Hadoop framework to parallel the association rules algorithm.The experimental results show that the complex network community detection algorithm proposed in this paper achieved high detection efficiency.