Exploring Public Discussions Regarding COVID-19 Vaccinations on Microblogs in China: Findings from Machine Learning Algorithms

Abstract

Large-scale, widespread COVID-19 vaccination is the most effective means of cutting off the spread of the novel coronavirus and establishing an immune barrier. Due to the large population base in China, it has been a very difficult task to establish such an immune barrier. Therefore, this study aims to explore the public’s discussions related to COVID-19 vaccinations on microblogs and to detect their sentiments toward COVID-19 vaccination so as to improve the vaccination rate in China. This study employed machine learning methods in the field of artificial intelligence to analyze mass data obtained from SinaWeibo. A total of 1,478,875 valid microblog texts were collected between December 2020 and June 2022, the results of which indicated that: (1) overall, negative texts (38.7%) slightly outweighed positive texts (36.1%); “Good” (63%) dominated positive texts, while “disgust” (44.6%) and “fear” (35.8%) dominated negative texts; (2) six overarching themes related to COVID-19 vaccination were identified: public trust in the Chinese government, changes in daily work and study, vaccine economy, international COVID-19 vaccination, the COVID-19 vaccine’s R&D, and COVID-19 vaccination for special groups. These themes and sentiments can clarify the public’s reactions to COVID-19 vaccination and help Chinese officials’ response to vaccine hesitancy. Furthermore, this study seeks to make up for the lack of focus on big data in public health and epidemiology research, and to provide novel insights for future studies.