Research On The Prediction Model Of User Behavior And Emotion Which Based On Sina Weibo

With the rapid development of the Internet,social network has become an important tool and platform for people to keep in touch and play.The core of the social network is the user,and the analysis and research of the user's behavior is the starting point of a deeper understanding of the operating mechanism of the social network.In China,sina weibo is widely accepted and used by the majority of users.Therefore,more and more experts and scholars have conducted research on sina weibo.Article mainly targeting weibo study,social network users against the related behavior and weibo affective forecasting problem is the current hot topic,using weibo API interface to get to the user's information in weibo,by use of Excel software to extract the data collection and analysis,concluded that the user has a certain similarity between forward and comments.Comprehensive support vector machine(SVM),K nearest neighbor(KNN)algorithm and CART decision tree for the calculation of a weighted average,was proposed based on hybrid classifier's weibo more emotional prediction model(hereinafter referred to as MCEPM),through the experiment of concrete operation and the SVM and KNN and CART weight coefficient calculation and contrast,MCEPM prediction effect and the weighting coefficient of the SVM and KNN and CART,elected to take the weight coefficient of combination,right through to the computing data in P accuracy and recall rate R and harmonic mean Fl on the comprehensive analysis,The MCEPM model can balance the prediction results of positive and negative samples of the three classifiers,and it can be concluded that the prediction capability of the model is relatively good.