Research On Identification Method And Application Of Influential Nodes In Social Network

Information spreading is inseparable from all kinds of networks,and important nodes will cause the phenomenon of information super dissemination,so the research on related scientific issues is particularly important.How to effectively identify and control important nodes,establish a comprehensive network management system,and create a clear cyberspace has attracted the attention of more scholars.In this paper,the K-core decomposition algorithm is optimized to realize the identification of important nodes,and it is applied to locate the propagation source point.Management measures are proposed for the important nodes and the propagation source point.In order to accurately identify the important nodes,it is necessary to compare several algorithms to determine the important nodes to ensure low time complexity and high utility.The K-core decomposition algorithm is a classical algorithm to judge the center node and analyze the node influence.However,there are some errors in the algorithm,which lead to the identification of important nodes including core like nodes.The characteristics of core like nodes are high k-core value and low propagation efficiency,which will reduce the accuracy of the algorithm.In view of this,based on the weighted network,the weight reset chaos algorithm is used to construct a zero model network to optimize the k-kernel decomposition algorithm in this paper,which can identify and exclude the core like nodes,and improve the accuracy of important nodes.Based on the simulation and the actual network,by comparing relationship of the average propagation efficiency between the similar-like core nodes and the same k-core layer nodes.It is verified whether the propagation capability of the similar-like nodes identified by the algorithm is lower than that of normal nodes with the same centrality.The SI model is used to simulate the information transmission,different transmission probabilities are set,and the average transmission efficiency is calculated by 10,000 times of repeated transmission.Experimental results show that,in the same time,the propagation efficiency of most core like group nodes is lower than that of true core nodes at the same level.In the application,the propagation source nodes can be located by important nodes.When the spreading source nodes user of a rumor has a certain network influence,the rumor is more likely to explode on a large scale.This paper assumes that the spreading source nodes user of a rumor is one of the important nodes,and the rumor has the characteristic of spreading to more important nodes in a shorter time.In the original network,the shortest paths between two important nodes are calculated,the connecting edges involved in each Dijkstra path are selected,and the sub-network is constructed together with the important nodes.The SI model is used to simulate the propagation,and the sum of the Dijkstra path length of I state node and all other important nodes is calculated.The I state node with the minimum Dijkstra path sum is selected as the propagation source node.Simulation and actual network are carried out to verify the effectiveness of the algorithm.The experimental results show that the algorithm can locate the propagation source node in the network with an accuracy of about 80%.In the study of information transmission,important nodes and transmission source nodes,as the key nodes in the network,are also the focus of social management.In this paper,management strategies are proposed for important nodes and propagation source nodes to suppress the spread of rumors.For important nodes,unified rules and regulations can be formulated for such users to prevent them from spreading rumors.When other rumors break out on a large scale,positive comments should be made in time to guide public opinion from these important nodes users.According to the characteristics of different source nodes,the source nodes can be analyzed and positioned,and then a series of management measures are put forward,including guiding the first users to make positive comments,punishing the source points,and publishing the means of punishment.