Research On Emotion Classification Of Chinese Micro Blog Based On Ensemble Learning

With the accelerated pace of economic globalization,the vigorous development of the Internet bring people all aspects of convenience.Microblogging text Contains a lot of information which related to people.The emotional analysis of microblogging text can promote product improvement,timely monitoring of public opinion,the correct implementation of the decision-making and implementation.This paper studies the emotional classification of Chinese microblogging which based on integration classifier.The article is divided into three categories:positive polarity,neutrality and negative polarity.First,the existing classical Chinese emotional dictionary is collected and sorted out.Then,it is difficult to identify the problem of network emotion words in microblogging emotion classification task.This algorithm is based on the Word2vec word vector model for the network emotion word polarity recognition algorithm,this algorithm from a small number of seed network emotional words to find a large number of new network emotional words.In order to reduce the risk of misclassification by considering the results of multiple classifiers,this paper designs an integrated classifier based on SVM,LR and RF for microblogging emotion analysis.In this paper,we introduce the network emotion words excavated in this paper to make the effect of emotion classification bigger and higher.Finally,we analyze the integrated classifier based on SVM.LR and RF The experimental results show that the introduction of microblogging emotion classification based on the integrated classifier constructed by SVM,LR and RF can improve the effect of emotion classification.