Identifying public concerns and reactions during the COVID-19 pandemic on Twitter: A text-mining analysis

Abstract

Efforts to control the current coronavirus disease 2019 (COVID-19) pandemic have led to national lockdowns around the world. Reactions to the rapidly evolving outbreak were shared on social media platforms. We conducted a mixed-methods analysis of tweets collected from May 10 to May 24, 2020, using MAXQDA software in conjunction with Twitters search API using the keywords: “COVID-19,” “coronavirus pandemic,” “Covid19,” “face masks,” and included terms such as “Queens,” “Bronx,” “New York.” A total of 7, 301 COVID-19-related tweets across the globe were analyzed. We used SAS Text Miner V.15.1 for descriptive text mining to uncover the primary topics in unstructured textual data. Content analysis of tweets revealed six themes: surveillance, prevention, treatments, testing and cure, symptoms and transmission, fear, and financial loss. Our study also demonstrates the feasibility of using Twitter to capture real-time data to assess the public's concerns and public health needs during the COVID-19 pandemic.