The Influence Maximization In Social Network

In recent years,with the improvement of Internet and Web2.0 technology,a variety of social networking services continue to emerge.People are increasingly accustomed to using online social networking platform to interaction and information release.Social network has therefor become an important medium for knowledge sharing,interactive communication and information dissemination.Influence maximization,as one of the key issues in the field of social network.It has many practical applications,such as the source of public opinion monitoring,the choice of agents in marketing and the positioning of water quality monitoring.Influence maximization is intended to find the most influential seed node set and how to find this set has proven to be NPHard.Existing method used to solve influence maximization is mainly focused on greedy algorithm,heuristic algorithm and community algorithm.But in the large-scale network to solve influence maximization,there still exist some problems,such as high time complexity,low precision and poor robustness.Therefore,it is a very important research topic to find an efficient way to solve the influence maximization in large-scale network.In view of the above problems,this paper studies the key issues such as network topology,influence propagation model,influence maximization algorithm and influence evaluation method.The main results are as follows.(1)To solve these problems,this paper proposes a hybrid heuristic algorithm named Degree and Influence Heuristic(DIH).According to the accumulation caused by using linear threshold model features,the computation of influence maximization is broken up into two stages: the degree discount heuristics and influence-based heuristics.We first employ the degree discount heuristics to active nodes and accumulate influence of nodes.Then employ the influence-based heuristics to look for the most influential nodes which will collect and outbreak in the influence of the first stage accumulation.(2)Based on the influence between nodes decreases as the distance increases and three degree principle of influence,this paper proposes a new method to evaluate the influence of the node.This method can calculate the global influence of nodes in the second stage of DIH algorithm at a faster rate.(3)In order to verify the validity of DIH algorithm,this paper compares it with several classical influence maximization algorithms in three real networks.The experimental results show that compared with the traditional heuristic algorithm,the algorithm can achieve better effect under the same efficiency and it has better robustness in the face of different network structures.