Research On The Influence Of Interactive Behavior Based On Interaction In Social Network

In recent years, with the development of social networks, the main approach to communication gradually changed from classical ways such as face to face review and letter and so on, to internet, which makes it very important to find the most influential K users in social networks, also known as social networks influence maximization problem. Influence maximization problem intends to mining the most influential collection of Top-k nodes in social networks. In previous study of this problem, people mostly find the most influential users just according to the topology of the network to, but ignore the frequency of interaction which is a very important factor to reflect the degree of closeness between the users. So that the excavation of users always have larger deviation with actual situation.For this consideration, we propose the user-interactive behavior-based influence maximization problem and build a UIB_IC influence propagation model based on user interactive behavior. In the UIB_IC model, for quantifying the extent of the interaction, this paper puts forward the concept of degree of interaction, summarizes the influence calculation method based on user interaction, and normalizes the data, by which we can calculate the activation probability between users. So, we propose GAUIB algorithm according to UIB_IC model. GAUIB algorithm, improved on greedy algorithm, determines the activation probability between users by the interaction behavior, which can measure the influence between users more accurately. In GAUIB algorithm, the accuracy can reach 63% through its sub module. In order to improve the computation efficiency of the algorithm, we integrate the algorithm of CELF to optimize the algorithm.Finally, this article, processing the data from tencent weibo, proves that GAUIB algorithm can find out the most influential set S of users based on user interaction behavior.