Research On The Influence Maximization Based On Community

The communication technology of influence in social network is a key to the application of marketing and opinion making.The spread of influence in the real world is often accompanied by two or more competitors.The maximization of competitive influence is to study how to choose the optimal initial seed nodes for each competitor.These seed nodes can ensure the maximum spread range of influence in the competitive environment.However,most of the existing researches fail to consider the role of inactive nodes in the process of information transmission,and the existing researches also neglect to consider the problem of community homogeneity,thus blocking the dissemination of information.Aiming at the above problems,this paper is to study a method to maximize the competitive influence based on community.The work of the thesis mainly includes the following contents:First of all,presents a new propagation model in the influence of competitive environment,the model not only considers the active users,and considers the non active users have less influence,which can help the competitor to find potential users that exist in the social network,thus ensure against in the competitive sales results.Then a community-based approach to maximize competitive influence is proposed.Firstly,the relevant concepts and problem descriptions of the competition influence maximization problem based on community are given,and then the community discovery algorithm is proposed to provide the basis for the subsequent influence maximization.Then,based on the boundary influence,a two-stage seed node selection algorithm is proposed,which is divided into heuristic stage and greedy stage.Finally,the time complexity of the algorithm is analyzed and discussed.This method solves three problems existing in the existing research: 1)It ignores the consideration of community structure,which leads to the slow speed of information transmission;2)The problem of information transmission obstruction caused by the homogeneity of the community;3)Low efficiency.Finally,design and complete the experiment.On three real data sets,such as facebook?combined,Wiki?Vote,and NetHEPT,comparative experiments of influence propagation model,experiment of influence law of different heuristic factors,comparison experiment of seed node selection algorithm,and seed node selection experiment were conducted.Experimental results prove the feasibility and accuracy of the proposed model and algorithm.