Analysis On Sentiment Orientation And Its Evolution Of Network Public Opinion With Text Mining

With the rapid development of the Internet,social network has become an important platform for users to obtain information,express their opinions and exchange views.Once the hot event occurs,online users can express their subjective information such as attitude,cognition,opinion and emotion to unspecific things or people with media like pictures,words and videos.Ways such as forwarding,review and praise facilitate the propagation of event.Meanwhile,if users add personal views or subjective emotions when they propagate information,then they promote the evolution of the event.In recent years,the number of network group events has risen sharply,causing a huge public response in Internet users.If the negative public opinion is not controlled and guided,public opinion will be very easy to extreme,and even endanger social security and stability.Therefore,it is necessary to study sentiment analysis of the users for online public opinion,and provide appropriate theoretical supports and countermeasures for the government to effectively master and monitor network public opinion events.This paper takes the “Luo Yixiao” hot topic event as an example,employed emotion analysis and tracked public opinion to analyze the public opinion information.The main research work includes: Firstly,collects event-related online public opinion data by the web crawler tools.Secondly,a comprehensive emotion classification lexicon is established after extends emotion words to HowNet emotion lexicon etc.At the same time,we can build sentiment orientation analysis model to identify and mark the polarity and intensity of emotion words.Thirdly,we can judge user's emotion type and statistical emotion frequency to dig and visualize analysis user's emotions.Finally,by using empirical analysis,we can divide the stage of public opinion evolution,and then analyze the characteristics and rules of user's emotion evolution.Meanwhile,it provides reference for the follow-up network public opinion guidance strategy.The experiments show that the network public opinion from generation to final demise is a complete life cycle,through the network of public opinion evolution of the scientific division phase,feature in various stages can be found:(1)In the beginning,the number of Network public opinion is small,and the attitude of Internet users to network public opinion events is complicated.However,the extraction of users' feelings in the text is helpful to further follow the development trend of the events.(2)The number of Network public opinion in the burst period is the highest,the user participation is the highest,the influence scope and the influence effect are very great.The attitude,opinion and emotion of the Internet users to the event can provide sufficient data base for the network public opinion analysis and monitoring.At the same time,the emotional tendency of the outbreak period largely defines the general trend of the emotional evolution of the network public opinion events,and more attention of the relevant departments should be pay to the evolution of the network public opinion emotion in the outbreak period,the public opinion is guided to develop in the right direction.(3)During the fermentation period,Internet users are more sensitive to new information and new developments in network public opinion events.Positive disclosure and public opinion disclosure can achieve good results at this stage.(4)Digestion and reflection during the period has low level of user participation,but reports is still needed to be tracked on the network public opinion,rumors is needed to avoided,the network environment is needed to be eliminated in order to avoid the secondary fermentation of the network public opinion.