Influence Maximization Algorithm Based On Topology Potential Research And Application

With the rapid development of microblog,wechat and other large-scale social networks,social networks have become an important platform for people to communicate with each other,and the connections and communication between users in these social platforms have become social networks.According to the principle of conformity,the influential user nodes can easily affect the ordinary user nodes.Therefore,the "social marketing" conducted by businesses with the help of social networks has developed rapidly.At present,there are some relatively mature models and algorithms for how to find the most influential user nodes.However,there are still some problems such as inaccurate influence propagation and low efficiency of algorithm operation,and how to effectively find the subsequent nodes after deleting nodes in social networks.(1)An algorithm of maximizing influence based on topological potential is proposed.The existing algorithms have the problems of inaccurate influence diffusion and low efficiency.In order to solve this problem,an algorithm is proposed.First,based on the topological potential theory,the nodes are classified;then,the heuristic method is used to form the candidate seed set;finally,the CELF(Cost-Effective Forward)algorithm is used to determine the optimal seed set.In addition,according to the Ca-AstroPh dataset provided by SNAP website,the influence diffusion range and algorithm running time are tested.Experiments show that the algorithm based on topological potential maximizes the influence propagation and improves the efficiency of the algorithm.(2)This paper proposes a selection algorithm of successor nodes based on simulated annealing.In the existing social network,there is a problem of how to find the successor node after deleting the node.Not all the seed nodes are willing to spread information.Therefore,it is very important to find a series of successor nodes to replace these useless seeds.In order to solve this problem,an algorithm is proposed.Firstly,the influence of the subsequent nodes is considered to reduce the economic loss caused by deleting seeds from the network for some reason.Then,select the maximum degree node and the neighbor of the deleted node as the candidate seeds.Finally,the simulated annealing algorithm is used to find the most influential seed set which is composed of subsequent seeds and original seeds.In addition,according to the Email-Enron data set provided by SNAP website,the influence diffusion range and algorithm running time are tested.Experimental data show that the algorithm ensures the accuracy of influence propagation and the efficiency of the algorithm,while reducing the loss caused by deleting nodes.(3)The application of influence maximization algorithm in Y company's marketing management system is realized.Using object-oriented and UML modeling technology,the system is designed,and key module model diagrams such as class diagram,architecture diagram,sequence diagram,database design diagram are given.The system displays some users with great influence,and the company selects marketing personnel to share the platform's beauty products to drive sales.