Improvement Of Influence Propagation Model Based On Dynamic Network

The Internet provides a variety of platforms for the dissemination of information,and the process of information dissemination is accompanied by the spread of influence.The analysis of the influence in the process of information dissemination can be applied to the following aspects.For example advertising,commodity marketing and public opinion control.The spread of influence in the network is analyzed by the influence propagation model.Because of the widespread use of influence propagation,there have been a lot of research on influence propagation models.In this paper,the key research of the influence propagation is study of the influence propagation model under the dynamic network.At present,most researches on the influence propagation model are carried out under static network.But in static network,above the network structure and the state of the node is invariable.And it will caused the limitation analysis of influence propagation that the results of the analysis were not consistent with the actual communication.The study of influence-spreading models in dynamic networks makes the propagation of influence more accurate.But there are several problems to be solved.First,how to determine the activation threshold.Second,the calculation of the activation probability.In this paper,the main work is improvement the calculation method of activation probability.Through the in-depth study of the existing model and the analysis of the influence factors of information dissemination in social network.Add sentiment to the calculation formula of activation probability.Then,analyzing the emotional transformation in the transmission process and get the sentiment conversion rate in the propagation process to improve the activation probability calculation formula.The improved activation probability formula is added to the independent cascade model to adapt to the dynamic changes of the network.The data is captured through sina weibo API,and include total of 15000 microblogs are to validate the improved independent cascade model.The evaluation index is used to evaluate the propagation effect of the model.Compared with the original model,the improved independent cascade model can predict the transmission of influence more effectively.