Desai, T., & Conjeevaram, A. (2021). The COVID-19 Infodemic: The complex task of elevating signal and eliminating noise. medRxiv. https://europepmc.org/article/ppr/ppr269180

University of California, Berkeley

Because of the influx of tweets about the COVID-19 pandemic on twitter, physicians created health-cares speciific microcommunities to be able to share research and information with each other. This paper studies the content of the tweets in these communities, categorizing them as well as looking at how much of it is noise. 

The COVID-19 Infodemic: The complex task of elevating signal and eliminating noise

Researchers looked at COVID-19 related tweets in 8 physician created communities on Twitter. between March 11 ,2020 and September 27, 2020. These were denoted by the hashtags #4rheum, #4MD, #4MDs, #2, #4MD, #4MDs, and #19Kidney. All of these hastags were either created on March 11 or March 12 of 2020 in order to create communities of physicians / medical professionals to share reliable research and information. 
These tweets (27071) were downloaded using an application coded in Java, stored, translated into English if not already, and then categorized into 5 domains (Diagnostics, Pathophysiology, Prevention, Symptoms, and Therapeutics) and 84 subdomains. Any tweet that could not be categorized was considered noise.

The COVID-19 Infodemic: The complex task of elevating signal and eliminating noise - methodology

- only 43% (9051) of the collected tweets were signal (able to be categorized into a domain and subdomain)
- as the pandemic progressed, there was a decrease in trend of signal tweets
-  domains with the most tweets were Prevention and Therapeutics
- even in these communities of medical professionals, there was still a lot of noise diluting information about the pandemic

Learn Before

Related