Research On The Problem Of Maximizing The Impact Of Social Network Blocking

Online social network is an important channel and carrier of information dissemination in today's society,forming a virtual society that interacts with the real world.Common online social networks are SINAWEIBO,TWTTER,FACEBOOK,etc.These vectors allow users to create their own accounts,post information,and spread messages.The information published by users contains some current social phenomena and hotspots,and topics often involve politics,entertainment,and economics.Therefore,the problem of maximizing impact has been widely applied in the fields of marketing,infectious disease suppression,public opinion monitoring,government policy promotion,and social stability.However,all of the above research work only studies the single spread in social networks.In fact,there are often different or completely opposite information,and opinions and innovative opinions are competitively spread in the same social network.This competitive impact spread may come from two competing companies,and in the market competition in which the two companies participate,they should try to attract people to pay attention to their own marketing activities.Therefore,the problem of maximizing the impact of the blockage is proposed to simulate the influence of the competitive relationship in real life.The existing blocking influence maximization algorithm ignores the influence of user preference on information dissemination,does not reach the maximum impact range,and runs for a long time.On the other hand,the existing dynamic social network blocking influence maximization algorithm directly utilizes the static social network computing method,but the social network is growing,and the static method of recalculating the most influential seed node every time the network is updated.Will generate a lot of computational overhead.In view of the two problems raised by the above blocking effect maximization,this paper has done research from the following two aspects:(1)Blocking effect maximization algorithm based on subject perception.The MTIB(Multiple Topics-aware Influence Blocking)algorithm is proposed to improve the impact range and reduce the running time.The main idea of the MTIB algorithm is to set the probability of the edge of the network topology to the propagation probability of the topic,and then calculate the seed set that maximizes the blocking effect.The MTIB algorithm generates a reverse weighting tree for each node.The nodes in the inverse weighting tree calculate a score that represents the blocking range of the node,and finally find the k nodes with the highest score as the blocking seed set.The experimental results of real social networks show that the MTIB algorithm has significantly reduced running time compared with the existing algorithms,and the impact range is greatly improved.(2)Blocking effect maximization algorithm based on dynamic social network.The existing dynamic social network blocking impact maximization algorithm utilizes the static social network computing method to generate a lot of unnecessary computational overhead.This paper proposes an incremental algorithm DBIM(Dynamic Blocking Influence Maximization)to reduce the calculation and use the pruning strategy to further reduce the running time.The DBIM algorithm first calculates the initial seed set in a static method in the initial network topology diagram,and at the next moment,updates the seed set according to the incremental part of the topology change and the seed set calculated at the initial time.The pruning strategy is used to further reduce computational overhead when updating seed collections.The experimental results show that compared with other algorithms,the DBIM algorithm has significantly reduced the time it takes to obtain a similar range of propagation.