Trend analysis of COVID-19 mis/disinformation narratives - A 3-year study

To tackle the COVID-19 infodemic, we selected as relevant for the analysis 42,410 articles from 460 unverified sources, that is, sources that were indicated by fact checkers and other disinformation experts as frequently spreading mis/disinformation, covering the period from 1 January 2020 to 31 December 2021. To identify the main narratives of COVID-19 mis/disinformation, develop a codebook, automate the process by training a classifier, and analyse the spread of narratives over time and across countries. We set up a process for classifying narratives of COVID-19 mis/disinformation in near real-time. The system relies on a multi-level codebook and a Transformer-based model. In the period from 1 January 2020 to 31 December 2021, we identified 12 supernarratives that evolved over time. With the availability of COVID-19 vaccines, a rise in anti-vax content occurred. We established a process that allows for daily monitoring of COVID-19 mis/disinformation. The experience gathered over the first two years of the pandemic will be useful to analyse mis/disinformation on other topics, such as climate change, migration, and geopolitical developments.

KOTSEVA Bonka;  VIANINI Irene;  NIKOLAIDIS Nikolaos;  FAGGIANI Nicolò;  POTAPOVA Kristina;  GASPARRO Caroline;  STEINER Yaniv;  SCORNAVACCHE Jessica;  JACQUET Guillaume;  DRAGU Vlad;  DELLA ROCCA Leonida;  BUCCI Stefano;  PODAVINI Aldo;  VERILE Marco;  MACMILLAN Charles;  LINGE Jens; 

2023-11-22

PUBLIC LIBRARY SCIENCE

JRC129010

1932-6203 (online),   

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0291423,    https://publications.jrc.ec.europa.eu/repository/handle/JRC129010,   

10.1371/journal.pone.0291423 (online),