Leveraging official government data for enhanced understanding of health crisis communication topics

Abstract

Taking full advantage of multiple official government data sources is crucial for government emergency management agencies and authorities to formulate scientific and feasible crisis communication policies and measures. The current study examined the case of the public health emergency to explore the association between government microblog activities and the evolution of the public health emergency. Moreover, the present research utilized latent Dirichlet allocation (LDA) topic modeling to detect dominant pandemic-related themes in the government microblog dataset and the government press release conference dataset. Research results indicate that the evolution of government microblog activities is generally similar to that of the pandemic outbreak. The current research also found that government microblog messages and press releases had consistent pandemic-related themes (medical treatments, nucleic acid testing, pandemic statistics, pandemic updates, press conferences, and material support). These results can assistant government agencies and stakeholders obtain a comprehensive understanding of the evolution of public health emergencies and public feelings, concerns, and issues, which will further improve their crisis communication ability.