Research On The Maximization Of Influence Of Node Sets In Social Networks

Mobile Internet,social networks have been integrated into people's daily lives increasingly,especially with the development of the Network social platform,more and more plays a very important role in the mode of social interaction and the way of information dissemination.In word-of-mouth marketing,merchants often select influential nodes as seed users,and through the seed nodes to spread the products of the merchants,so that more consumers know and buy the products of the merchants.Mapping the marketing model to the network environment enables the seed information to be maximized by the seed node.In social networks,the influence of seed nodes plays an important role in the control of network public opinion and the outbreak of rumors.Therefore,in the social network environment,the research on the maximization of the influence of nodes has great theoretical and practical significan-ce.Influence maximizing in social network is usually studied from two aspects of algorithm and model.This paper improves the existing work through in-depth analysis of the research results of this problem in a few years,and verifies the effectiveness of the proposed algorithm through real data sets.The research results of this paper are mainly reflected in the following aspects:?1?In order to study the influence of subgroup structure on information diffusion in the network,this paper proposes a k-core refining algorithm based on controlling the diffusion cost.The algorithm performs a range refinement on the k_max-core for the node neighbor overlap analysis of the k_max-core and(k_max-1)-core.The algorithm solves the problem of k_max-nuclear influence overlap,and further narrows the range of influential nodes in the k_max-core.Under the control of diffusion costs,as far as possible to ensure the spread to a larger range,increase the applicability to practical problems.Through simulation experiments and microblogging empirical analysis,the results show that the algorithm has good effect.?2?Aiming at the problem that the end nodes in the network have no obvious effect on information dissemination,this paper proposes a method based on two-stage generation of candidate node sets.This method incorporates influential nodes into the candidate system as much as possible through the node primary selection and the node supplementation.?3?In order to solve the limitation of node single attribute evaluation node influence,this paper proposes a hybrid attribute strategy evaluation node algorithm?NCBI?.The NCBI algorithm comprehensively considers the influence of the neighbors of the node and the dual criteria of the influence of the node itself.The comprehensive value of the NCI index,the UBC index and the CC index is used as the criterion for judging the influence of the node.The algorithm solves the one-sidedness of the importance of a single attribute evaluation node and increases the joint evaluation of node attributes.It is beneficial to reduce the error of node impact assessment and is more suitable for large-scale networks.The experiment selects the actual network to verify the algorithm in the independent cascade model and compares it with the related algorithm.The results show that the algorithm has good effectiveness.