| The emergence of social networks has provided people with new media information,and it has also greatly increased the possibility of outbreaks of online public opinion,and online public opinion has great impact on real social security.With social networks as the medium,relevant reports on various public emergencies will spread on social networks in a very short period of time.The social sentiment expressed by the Internet public opinion mainly has two types of emotional expressions: positive emotion and negative emotion.Especially the negative emotion expressed by online public opinion has serious impact on social security,so the emotion expressed by online public opinion Excavation and supervision are essential.The specific work is as follows:Firstly,a framework analysis of emergent network public opinion evolution based on text sentiment analysis and implicit topic mining is constructed,which involves Chinese word segmentation based on hidden Markov model,word vector construction model based on deep neural network,based on support vector machine.Text sentiment analysis and text topic mining based on topic models.Secondly,based on the sentiment analysis framework,this paper proposes a method for analyzing the esoteric theme evolution of emergencies in time series.The study includes the evolution of the sentimental polarity of emergency network lyrics,the evolution of sensational themes in the process of dissemination and diffusion,and the introduction of time-slicing into the topic model,and then analyzes the law of public sentiment polarity and lyric theme.Finally,the dissemination network and the subject evolution empirical analysis of each stage of the specific life cycle of the emergent event are realized,and the hidden topics in the process of the propagation and propagation of the emergency are mined.Study the causes of the theme changes in the evolution of emergencies,and verify the role of sentiment analysis in the evolution of online public opinion. |
PDF Full Text Request