Research On Information Diffusion And Control In Location-Based Social Networks

In recent years,with the rapid development of the Internet,online social networks have gained significant popularity,the related problem of social networks has been hot issues.The problem of information diffusion and control in online social networks has been extensively studied.However,in mainly recent works,information diffusion in the physical world is not considered.In our daily life,words of mouth among people in the physical world is an important way to spread information.It asks for a new requirement: we need users' online social relationships and location information.Location-based social networks as emerging social networks can meet requirement that we need users' online social relationships and location information.Location-based social networks can provide users' online social relationships and location information.We build a dual-layer network structure model consisting of social relationships and location information in location-based social networks and then achieve the purpose of message dissemination and control in the physical world and online social networks.The lower layer is composed of geographical information of users in the physical world.The upper layer is composed of social relationships among users in online social network.In the process of information diffusion,we choose some user nodes as seed nodes to spread information to other nodes to make information diffusion faster and more widely.In the process of information control,we start to control information diffusion from these selected seed nodes to avoid information spread further.Next,we describer the way to spread information and diffusion model.Afterwards,a heuristic algorithm for selecting a set of seed nodes with selecting an initial seed node and a non-initial seed node is given.In the experimental part,we used two real datasets for simulation experiments.We conducted experiments on single-hop and multi-hop information diffusion modes in the double-layer network structure model and each single-layer structure model,and we compared the heuristic algorithm we proposed with the greedy algorithm and the random selection algorithm,then we tested the parameters in the experiment.The final experimental results show that the speed and breadth of information diffusion in our dual-layer network structure are significantly better than those in any single layer.The experimental results of the heuristic algorithm we proposed are obviously better than other algorithms.The experimental parameters we tested had a significant effect on the experimental results,information diffusion was controlled obviously.This shows that our proposed model and heuristic algorithm are all valid.