A New Internet Public Opinion Evaluation Model

Sheng-Tsung Tu, Louis Lu, Chih-Hung Hsieh, Chia-Yu Wu

OAI: oai:igi-global.com:287603 • DOI: 10.4018/IJBDAH.287603

This research retrieved public opinions on the novel coronavirus pandemic with the aid of the DiVoMiner. The data were collected by setting keywords via qualitative comparative analysis (QCA) and automated computational approach, and the collected data were analyzed subsequently. The present study divided keyword collections into three categories, namely the name of diseases, government policies and COVID-19 events. It was found the retrieved Internet public opinions on COVID-19 was the largest in number and contained the least noise when the three categories of keywords appeared at the same time. Therefore, the data of Internet public opinions = the name of diseases × (government policies + COVID-19 events). This research found that an event that happens daily will affect the number of Internet public opinions on social media and forums after it has been reported. The strong negative emotion conveyed through the Internet public opinion may turn into a positive one if the event is dealt with properly after positive focus words represent the same proportion as negative ones.