The data collected consists of mostly irrelevant Tweets and simply not relevant to the misconceptions of interest. However, since Covid-19 has not come to an end yet, this data is still changing and can be completely different now 

University of Michigan - Ann Arbor

- Studied 4,790 covid-19-related Tweets from March to April of 2020 
- Used Wikipedia to determine 86 of the common misconceptions regarding covid-19
- Used BERTScore to calculate the similarity of Tweets to common misconceptions
- 100 of the most similar Tweets were then studied in detail
- Categorized each of the 100 Twitter misconceptions as misinformative, informative, or irrelevant. Informative means the post was simply trying to correct the false information presented in other posts

Detecting COVID-19 Misinformation on Social Media Data Collection

Hossain, T., Logan IV, R. L., Ugarte, A., Matsubara, Y., Singh, S., & Young, S. (2020). Detecting covid-19 misinformation on social media. https://openreview.net/forum?id=tRDw11nID9 

Learn Before

Related