Research On Topic-based Impact Maximization Algorithms In Social Networks

In recent years,with the social network becomes more and more popular,the research of influence propagate has become a hot spot.The main areas of influence propagate are the following areas,the impact of communication modeling research,learning the impact of transmission probability,the impact of the maximization of research.Influence maximization problem refers to the dissemination of a commodity in the social network,due to the limited budget,can only choose k nodes as a set,which is initially given a free trial of the commodity users,K users can reach the expected effect in the scope of the largest social network,which can spread through social networks not only can be a commodity,but also news,movies and even rumors.In this paper,we mainly study the problem of maximizing the impact of the theme:First of all,the local maximum effect,propagation only consider a single th eme,but in the real effect in the process of communication for each propagation has a different theme distribution,based on local influence maximization of the considered factors that influence the scope of the subject,the calculation is mor e accurate,this paper proposes a method of calculation local influence based on theme,and the theme of the design of local influence maximization algorithm ba sed on this method.The experimental results on several real datasets show that t he topic is more accurate in the calculation of the influence degree.Secondly,the influence of media reality not only considers the propagation on the theme distribution,also need to consider the interest distribution of the us ers themselves,the existing model in calculating the impact probability,or did n ot consider the subject distribution,or consider propagation on the subject of dist ribution,and not taking into account the interest distribution and subject distribut ion propagation of users in this paper,effects of interest spread based on the mo del,and puts forward the influence of interest maximization algorithm based on on the basis of the model,compared with the traditional propagation model com parison,this model in dealing with the influence maximization problem is more accurate.Finally,considering the effect of geographic factors to influence the basic factors in the theme,because in many real-world social network applications,geographical factors play a key role,we proposed the influence interest maximization problem based on geographical location awareness,in order to improve the efficiency of the algorithm,the selection of design the candidate set and calculate the upper bound of two algorithms.Experiments on several datasets demonstrate the effectiveness and efficiency of the proposed algorithm.