Design And Implementation Of Sentiment Analysis And Prediction System Based On Rednet

Nowadays,the Internet has been spread over every corner of people's lives.More and more people are participating in Internet commentary and are expressing their opinions in forums,shopping website comment areas and other places.However,because the openness of the Internet has bred rumors to a certain extent,the speed of information transmission on the Internet has also accelerated the spread of rumors,which has caused great harm to the stability of society.Therefore,the analysis,prediction and monitoring of Internet public opinion are particularly important.This system aims to help the government understand the public opinion of the Internet and understand the changes in the public opinion after the implementation of the policy.At the same time,it has provided technical support for the better regulation of Internet public opinion and further improvement of government management capabilities.This system takes the red network hot topics and topic review information as research objects.Based on the research of text sentiment analysis technology and current topic trend forecasting,a network public opinion analysis system with public opinion monitoring and trend forecasting function is designed and implemented.The main research contents of this article include:Firstly,the research background and significance of the system are expounded.The research status of text sentiment analysis in China and abroad is introduced.The key technologies and theoretical knowledge needed to implement this system are summarized.And analyze the functional requirements of the system.Secondly,according to the demand analysis,the design scheme of the system is determined and implemented.It is mainly divided into three modules: page information collection,sentiment trend analysis,and public opinion evolution trend forecasting.Get relevant data and build an index by using web crawler technology.Semantic-based sentiment analysis technology is used to analyze and monitor the positive emotions,negative emotions,and neutral attitudes that the Internet users hold on topics,and to use visual techniques to present relevant emotions in a graphical form.Finally,this system adopts the heat value calculation method that the present scholar puts forward to consider the topic click rate,reply rate and sentiment tendency score and uses the gamma model to predict the evolution trend of public opinion,andverifies and realizes the idea.At the end of this article,we have compiled monthly statistical data on the incident of the “ New Garbage Incineration Plant in Xiangtan Jiuhua Development Zone”.We have drawn scatter plots and line graphs through the form obtained from its situation prediction module,and found out the system's topic development curve and topic realities.The evolutionary trends in China are roughly the same,and have successfully completed the forecasting function of the evolutionary trend of public opinion.