Modeling And Application Research Of Users Interaction Behavior In Physical-social Binary Space

With the continuous promotion of mobile network technology(4G/5G)and the increasing popularity of portable intelligent terminal devices,various activities and behaviors of human beings in daily life generate a large number of user behavior data.The massive user data provides the possibility for the comprehensive disclosure and understanding of user behavior rules.On the one hand,everyone lives in the real physical world,and any movement will leave traces in the physical world.On the other hand,human beings are called "social animals" and inevitably communicate with people.To explore the laws of human activities in physical space and interactive behaviors in social space makes human have a more comprehensive and thorough understanding of themselves.Starting from the user interaction behavior in the physical-social binary space,this paper studies the mobile behavior preference and social network characteristics of users,mainly including three parts:(1)Analyze the user's residence time in different locations and other characteristics through the geographic location information in massive historical data,describe the user's preference for different location areas,excavate the information rule of user check-in Point of Interest(POI)and represent the user's preference for different location types;(2)Build social networks based on friends' information among users,and analyze the centrality of degree,eigenvector,small-world characteristics and other social network characteristics;(3)Combined with physical and social binary spatial information,using the spatial location in the physical world,POI preferences and social network three aspects,to build dual space together with physical and social characteristics of the user interaction model,represent the strength of social relationship between users,and conduct experimental analysis and verification on the real data set,The results show that the constructed geographic social network can effectively divide the originally complex social network according to the strength of the social relationship between users.Considering the physical-social duality of user interaction,the above model is applied to mobile crowd sensing user recruitment.Based on the two information dissemination models,the propagation process of task information in social networks is analyzed and predicted;The utility evaluation mechanism and task incentive mechanism were constructed to calculate the execution utility value of users recruited by different mobile crowd sensing tasks and the total task cost;The optimized algorithm is used to select the best recruiting user set;Finally,experimental verification is carried out on the real data set.The results show that the method adopted in this paper not only ensures the high completion rate of mobile crowd sensing task,but also significantly improves the response speed of mobile crowd sensing task and user resource utilization rate.