Dynamic Diffusion Community Detection Algorithms Based On Central Nodes

Community detection is one of the key research directions in the study of complex networks,and the main purpose of community detection algorithm is to accurately and efficiently identify community structure from complex networks.This paper proposes a community detection algorithm based on central nodes,which first finds the key nodes in the network,and then expands the labels of the key nodes,the specific work is as follows:Most of the existing community detection work is based on the static topology information of the network,ignoring the dynamic interaction on the network,this paper focuses on the dynamic process on the network and proposes an overlapping community detection algorithm based on nonbacktracking matrix.Firstly,in order to avoid ignoring small degree nodes,a node centrality evaluation index based on the number of nonbacktracking paths is proposed.Secondly,in order to model the multi-scale social interaction mode occurring on the network,a new edge membership vector is found to represent the community belonging of the node,the central node is linked with the community detection,and the dynamic system is used to represent the dynamic distribution process of community members,and the overlapping community detcetion in the network is completed in two steps.There are some problems when applying the density peak clustering algorithm directly to community detection,such as the need for manual intervention to select clustering centers,and this paper proposes a nonoverlapping community detection algorithm based on the leader node.Firstly,this paper continues to use the nodal centrality evaluation index based on the number of non-backtracking paths,and uses Chebyshev’s inequality to select potential community centers with higher density and large distance,and treats each community center as a separate community and constitutes the initial community structure.Secondly,a new membership vector is proposed to represent the community belonging of the node,and the community detection in the network is completed in two steps.Finally,in order to verify the effectiveness of the two methods,they are applied to real networks and artificial networks and compared with the classical community detection algorithm,and the experimental results show that the proposed method has great advantages in detection accuracy,and can more accurately detect the community structure in complex networks.