Information Dissemination And Prediction Algorithms Of Social Network

In recent years,with the promotion and development of social networks,people's communication methods have become more flexible and convenient,and communication between different regions is no longer limited by distance,which reduces the cost of maintaining human relationships greatly.Compared with the traditional media,social network has more users,and the speed of information dissemination is more advanced by the order of magnitude,which make social network becomes the mainstream media of the information age.However,a large amount of information is spreading in social networks with unprecedented speed and breadth.On the one hand,it brings convenience and opportunities for people's life.On the other hand,it also spread some false rumors.Therefore,it has important theoretical significance and practical value to research the process and the rules of information dissemination in social networks so as to regulate and use social network platforms more rationally and effectively.This thesis focuses on the information propagation models and hot trend prediction algorithms of social networks.The main contents are as follows:(1)For the traditional information propagation model is hard to reflect the rules of information dissemination in social networks accurately,this thesis studies existing social network information propagation models and proposes an online social network information propagation model,which based on node status.This model adds new node states and new transition state between nodes in the epidemic model.Compared with other models,it is verified that our model can reflect the information dissemination rules of real social networks more realistically.(2)In order to improve the prediction accuracy of current hot trend prediction algorithms on social network,this thesis proposes a social network hot trend prediction algorithm based on LSTM.Firstly,a prediction model is constructed according to the selected data set.Then,we propose the framework of social network hot trend prediction algorithm,which to describe the specific implementation of the prediction algorithm.In order to get the optimum results,the related parameters of the algorithm are further evaluated and selected.Finally,the accuracy of the algorithm's prediction was analyzed by comparing experiments.Experimental results show that the social network hot trend prediction algorithm based on LSTM has better performance in data prediction.In summary,this thesis builds an online social network information propagation model,and proposes a social network hot trend prediction algorithm based on LSTM,which improves the prediction accuracy of hot trend prediction algorithms on social network.