| description abstract | Due to its near-real-time crowdsourcing nature, social media demonstrates a great potential of rapidly reflecting public opinion during emergency events. However, systematic approaches are still desired to perceive public opinion in a rapid and reliable manner through social media. This research proposes two quantitative metrics—the fraction of event-related tweets (FET) and the net positive sentiment (NPS)—to examine the intensity and direction dimensions of public opinion. While FET is modeled through normalizing population size differences, NPS is modeled through a Bayesian-based method to incorporate uncertainty from social media information. To illustrate the feasibility and applicability of the proposed FET and NPS, we studied public opinion on society reopening amid COVID-19 for the entire United States and four individual states (i.e., California, New York, Texas, and Florida). The reflected trends of public opinion have been supported by the reopening policy timeline, the number of COVID-19 cases, and the economy characteristics. This research is expected to assist policy makers in obtaining a prompt understanding of public opinion from the intensity and direction dimensions, thereby facilitating timely and responsive policy making in emergency events. | |
