5.2 Utilizing natural language processing to analyze and understand the online public’s opinions and concerns towards COVID-19 
Along with the coronavirus pandemic, another crisis appeared in the form of large-scale fear and panic, and incomplete and often inaccurate information aggravated the phenomenon. Therefore, there is an urgent need to solve and better understand the information crisis of the new coronavirus pandemic and to measure public sentiment in order to implement appropriate information transmission and policy decisions.
Using natural language processing, researchers can monitor popular online comment boards to better understand public concerns during the novel coronavirus pandemic.
Public health officials can use natural language manipulation technology to track the topic of COVID-19 that has surged in interest on online forums. These insights can help policy makers understand public health/organizational concerns and priorities, while reducing the spread of misinformation. Real-time analysis of public response can make people aware of changes in public priorities, fluctuations in health conditions, and the adoption of public health measures, all of which will have an impact on individual and group levels of health.
Story 14: COVID-19 Public Sentiment Insights and Machine Learning for Tweets Classification
How people feel during this pandemic? This is a very complicated task as billions of people exist worldwide. However, nothing is impossible with AI. On June 11, 2020, a study by the University of Charleston determined public sentiment related to the pandemic through the Twitter and internal resistance health statistics software of the Coronavirus Group Health Organization, and sentiment analysis software. Through the use of descriptive text analysis and necessary text data visualization technology, they have proved that over time, as the new coronavirus pneumonia in the United States approaches its peak, the development of fear also reaches its peak (see Figures 31 and 32).
