Research On The Local Community Detection Algorithm Of Social Network

The impact of the rapid development of the mobile Internet has gradually penetrated into all levels of society,accelerating the integration of real interpersonal relationships and virtual social networks.Social networks help people establish more direct connections,facilitate communication,and expand communication circles.At the same time,social networks also provide convenience for criminals.Criminal gangs often use social network framework contact channels on social platforms to facilitate their criminal activities.Research on social networks,local community detection based on target users can play a significant role in assisting casefinding personnel to quickly discover criminal gangs,dig hidden clues to detect crimes,investigate criminal suspects and discover the core groups of criminal gangs.It has important theoretical and application value.This article analyzes the formation of local communities of target users by studying the structural characteristics of social networks and combining graph theory,mainly including the following two aspects:(1)Propose a local community detection algorithm based on specific community structure characteristics.The algorithm considers the structural characteristics of the social network and the attributes of the nodes in the social network.Combined with the theory of graph traversal and the shortest path,it can be divided into 5 algorithms: G2 G,G2P,P2 P,P2G,and Coregroup for specific situations to meet the needs of the public security department.With the help of the CKGraph graph database’s powerful analysis and calculation capabilities,the algorithm can obtain seconds or even milliseconds of response and can assist the case-handling personnel to quickly find the local community where the target user is located,or even find the core group in the local community,which has high practical application value.(2)Construct an algorithm visualization prototype system.The front end of the system is developed based on the classic modular framework Bootstrap.The visualization of the relationship diagram uses the Echarts data visualization component.The system backend is developed based on Python’s Django framework,including CKGraph graph database,logic control module,data preprocessing module,and algorithm module.The logic control module realizes the process control of the algorithm execution.The data preprocessing module preprocesses the data in this article,including extracting the social network relationship of thetarget node and the label attribute of the node.The algorithm module is the specific algorithm involved in this article,including G2 G,G2P,P2 P,P2G,Coregroup 5 algorithms.