Monitoring the internet public opinion of mountain flood disaster using big data analysis

Abstract

The mountain torrent disasters have caused severe damage and loss of life and property to the society and the economy, which attract the public’s attention. Timely monitoring the characteristics and evolution of internet public opinion in mountain torrents is of significant impact on the emergency response and management. The primary objective of the paper is to analyze the time and spatial characteristics, attention key points, emotional trend and participants of the public opinion in mountain torrent disasters using big data and statistical analysis. The current situation is investigated and a prediction method using correlation analysis is proposed, which provides reference for decision makers to formulate response plans according to public opinion from internet. The data mining method is used to obtain relevant data based on microblog API interface, and the data is processed by statistical methods. Combined with the content analysis method, the characteristics of public opinion in mountain torrents are comprehensively analyzed, and the hotspots of public concern and the trends of public opinion are obtained. The results show that the eruption time of the public opinion coincides with the disaster occurrence time. From spatial characteristics analysis, the flood disasters in China mainly occur in the North China Plain, the middle and lower reaches of the Yangtze River, the Pearl River Basin, the Sichuan Basin, and the Northeast Plain. The public opinion attention degrees are highly positive correlated with the occurrence of mountain torrent disasters, and the correlation coefficient is 0.96. Citizens not only pay attention to the disaster situation and rescue work, but also focus on the improvement of the disaster prevention work. Public emotional trend is mainly based on expressing hope for a safe life, pity and sympathy for the victims, panic about the severity of the disaster, objective evaluation of the situation, as well as admiration and concern for rescuers, and suggestions on early warning education. As an opinion leader, People Daily has great influence on the internet public opinion. At different stages, participants should work together to maintain a healthy public opinion environment.