Network Public Opinion Simulation of Public Health Emergencies Based on LSTM Model

Abstract

Based on the analysis of the development of public opinion and the trend of public opinion enthusiasm rapid changes, the root causes and evolution laws of public opinion on public health emergencies were revealed, and the active guidance mode of public opinion on public emergencies was proposed, which can provide solutions for the government to find public opinion guidance strategies and public opinion regulation. This paper uses web crawler technology to crawl the public opinion data of “Xi’an High tech Hospital Pregnant Women Abortion Event” from January 3, 2022 to February 28, 2022, uses Baidu Index as a measure of online public opinion enthusiasm, and uses multiple indicators such as blog posts, forwarding, comments, etc. to propose a prediction model of online public opinion enthusiasm for public health emergencies based on LSTM neural network, The prediction simulation of public health emergency public opinion based on Sina Weibo data was completed. The results show that LSTM series models have good fitting effect on the public opinion heat data, the optimized LSTM model can be used to accurately control the development of network public opinion, which is conducive to the guidance and governance of network public opinion and the initiative of public opinion control. This paper provides reference for public opinion prediction and guidance of similar public health emergencies.