Research On Influence Maximization Algorithm Of Online Social Networks

With the increasing variety of social networks,people's demands for social networks are also increasing.Influence-maximizing nodes,as network focus and dissemination source nodes,play a key role in people's daily social information decision-making.This thesis is devoted to analyzing the influence maximization problem in social networks,mainly focusing on local node propagation and influence range values,considering local node optimization and global diffusion respectively,and solving the influence maximization problem when social networks are based on target perception.The specific research contents are as follows:(1)The influence of local nodes and the overall diffusion are studied.Aiming at the fact that the social network influence maximization problem cannot satisfy the large-scale diffusion of nodes,propagation range,and time efficiency at the same time,a new node influence maximization algorithm is proposed based on the idea of local node optimization and degree discount.For optimization,the value of local node influence is calculated by constructing the NAV function.Secondly,a node activation algorithm is proposed,combined with the idea of degree discount to filter candidate nodes;the DMAP function is constructed to use the filtered candidate nodes for global diffusion.It is 11.3% higher than the traditional degree discount algorithm;it is four orders of magnitude faster than the traditional degree discount algorithm in terms of time efficiency.(2)Aiming at the problem of target perception and distributed incentive value query,a UAIA algorithm is proposed to estimate the influence relationship between users according to the historical behavior of each pair of users,so as to identify influential users in unknown social networks.An adaptive allocation incentive algorithm is proposed to determine the incentive value,and the target-aware RR index structure is used to effectively extract the user's behavior information in the online social network,and quickly query the target user and the user's overall status in the network according to the query model.Finally,four datasets and one synthetic dataset are used to evaluate the performance of the proposed method.The experimental results show that the proposed algorithm can not only obtain high-precision query results and influence spread,but also has high query efficiency.To sum up,the article analyzes the dissemination range,running time and influence user query influence range of online social network influence maximization from different angles.Model,and carried out theoretical and experimental analysis on these two models,the relevant conclusions obtained provide a strong theoretical basis for the study.