| The development of science and technology makes it no longer a dream for people to access the Internet anytime and anywhere and communicate across time and space.The emergence and popularity of various social platforms have expanded the communication channels for users.Different from traditional face-to-face conversations,instant information sharing across regions has become the mainstream choice for contemporary users.Netizens express their opinions on these platforms through operations such as forwarding,commenting,and liking.These information containing the true feelings of the publishers are easily aggregated into online public opinion after being disseminated on the Internet.Once the development of network public opinion exceeds a reasonable and controllable range,it will cause the dilemma of public opinion governance,affect the governance of cyberspace,and threaten social stability.This paper uses data mining as a means,combined with spatial statistical methods to supplement the time evolution of user emotions,and uses spatiotemporal analysis methods to better provide reference value for the reasonable control and guidance of public opinion.First of all,the article determines to take the "Henan Rainstorm" incident as a case,and conduct experiments by crawling the Weibo content,Weibo comments and commenters’ personal information published by the official Weibo on Sina Weibo as the original data.Secondly,in the time series analysis,according to the Pearson coefficient,the standard for determining the stages is the amount of Weibo comments,and the public opinion cycle is determined into three stages: the outbreak period,the volatility recession period,and the extinction period.Thirdly,the time series analysis of online public opinion is realized from the user emotions and hot topics,the distribution characteristics of user attention are explored from the number distribution of user comments,and whether there is an obvious public opinion emotion aggregation space from the distribution of user comments sentiment value,according to these three aspects.The temporal and spatial characteristics of the emotional situation evolution of network public opinion are used to interpret the internal mechanism of the evolution of public opinion in order to obtain general laws.The research found that:(1)The evolution of user emotions,the evolution of public opinion topics and the distribution of user emotions in the "Henan Rainstorm" event had obvious stages.The user’s emotional tendency is generally dominated by positive emotions,but at special time points,there are cases where the average value of emotions and the proportion of positive emotions drop to the lowest point.(2)The evolution of public opinion topics in different stages shows obvious thematic correlation laws,which are continuous and inherited.(3)From the perspective of space,the distribution of user emotions has global coverage and participation of the whole people,and has obvious aggregation characteristics.It is manifested that there are obvious user emotional cold spots in the northwest of China,and emotional hot spots are mostly located in the eastern region.Based on the research conclusions,this paper proposes that online public opinion guidance can be carried out from two aspects: the main body of communication and the characteristics of temporal and spatial changes,providing auxiliary decision support and basis for relevant management departments in public opinion research and judgment,analysis and prediction,and reasonable guidance. |