Research And Prediction Of User Centered Micro-blog Information Forwarding

Micro-blog information forwarding has become one of the key issues in the research of information dissemination.The forwarding probability and propagation trend of predictive information have important application value in information dissemination,public opinion monitoring product recommendation and so on.The existing research is mainly based on the network structure and the history of information dissemination rules to predict the future trend of information dissemination,ignoring the individual differences between users.In the forwarding prediction based on user behavior,the main factor is to study the factors that are forwarded from the perspective of the information publisher.In this paper,the user centered,standing in the perspective of the recipient of information,through the mining of the main factors affecting the user's forwarding and machine learning in the classification algorithm for prediction,the main work is as follows.First of all,according to the needs of the problem through the platform provided by the API crawl research data sets,including user information,micro-blog information,user relationship information and forwarding relationship information.We analyze and describe the characteristics and integrity of the data set,and then complete the preprocessing of the data set according to the results of the analysis.Then,we explore the important features that affect the user forwarding,including the character of the information publisher,the characteristics of the receiver and the interaction between users.The characteristics and affects of the selected features are analyzed by mining the relationship between feature and forwarding.Finally,using Support Vector Regression,Naive Bayes and Random Forest three classification algorithms and combined with the impact factors of information forwarding,this paper forecasts the user's forwarding information.Based on the experimental results,the best classification algorithm is selected to simulate the real forwarding process.Through the analysis of the model,it is proved that it is necessary to study the characteristic of the user's information to predict the behavior of information forwarding,and the importance of different factors affecting the behavior of information forwarding is obtained according to the ROC curve.