Sentiment Analysis Of Weibo Based On Expand-dictionary And Rule

Social networking has developed rapidly in recent years.As a social platform,Weibo stands out from many social platforms because it has the characteristics of immediacy,interactivity,and convenience.More and more people get information,share their feelings,and exchange views through Weibo.Through the emotional analysis of the information in Weibo text,we can learn about the opinions of the masses,understand the attitude of the users,and learn the information of the products.This is of great benefit to the government's policy promulgation,the orientation of the enterprise and the improvement of merchants' products.Weibo has become an increasingly important social tool for people,and emotional analysis of Weibo text has received more and more attention of experts and scholars.The main task of the Weibo's emotion analysis is to judge the emotional inclination.The emotion tendency is divided into three categories:positive,negative and neutral.Weibo text is different from the traditional text in the content and format.When analyzing the emotions of Weibo text,it is necessary to add the consideration to the characteristics of Weibo on the basis of analyzing the emotion of traditional texts.In this paper,the existing basic emotional dictionary is analyzed and integrated,and the network emotional dictionary is created to supplement it.Aiming at Weibo text,we create emotional dictionary in the field of Weibo and Weibo expression dictionary to expand the emotion dictionary.The semantic rules of text are analyzed and the corresponding weights are set up.Finally,the emotional tendencies of Weibo text are analyzed based on Extended Dictionaries and Semantic Rules.The main contents of this paper are as follows:(1)Emotional analysis of texts is based on the sentiment dictionary.The existing resources of emotional dictionary are organized and analyzed,and a comprehensive basic emotion dictionary is created according to the voting rules and the priority rules.Then this paper analyzes the users' habits on the Internet,and selects the emotional words in the lexicon of popular input to create the online emotion dictionary.(2)Creating an emotional dictionary in the Weibo field to supplement the traditional emotional dictionary.Based on Weibo corpus,the extended SO-PMI algorithm is used to create the emotional dictionary of Weibo automatically,and the selection method of benchmark words and candidate words in the process of creation are improved.Then we select the commonly used expressions of Weibo according to the word frequency to create a Weibo expression dictionary.(3)Semantic rules affect the emotional analysis of text.Semantic rules are analyzed from the two aspects of words and sentences.The rules of words mainly analyze the modified rules of negative words and degree words,and the rules of sentences mainly analyze the sentence patterns rules and the rules between sentences.In addition,the rules of the end sentence of text and the weighted rules of text and expression are also analyzed.Finally,in order to validate the effectiveness of the method proposed in this paper for sentiment analysis of Weibo based on extended sentiment lexicon and semantic rules,the COAE2014 Weibo test data are used to carry out the experiment.The experimental results show that compared with the traditional method of judging the emotion of Weibo,the improved method proposed in this paper improves the accuracy of the emotion analysis of Weibo.