A Second Pandemic: Fake News on COVID-19 Vaccines

With the emergence of the COVID-19 pandemic, the political and the medical aspects of disinformation merged as the problem got elevated to a whole new level to become the first global infodemic. Fighting this infodemic is currently ranked very high on the list of priorities of the World Health Organization (WHO), with dangers ranging from promoting fake cures, rumors, and conspiracy theories to spreading xenophobia and panic. Addressing the issue requires solving several challenging problems such as identifying messages containing verifiable claims, determining their check-worthiness and factuality, and their potential to do harm as well as the nature of that harm, to mention just a few. 

With this in mind, a team of scientists from the Qatar Computing Research Institute (QCRI), part of Hamad Bin Khalifa University (HBKU) explored the parallel pandemic of mistrust and how it may hinder our collective response to this disease. This effort is part of QCRI’s mega-project Tanbih, which aims to limit the impact of “fake news”, propaganda, and media bias. The team is led by Dr. Preslav Nakov, principal scientist at QCRI. The core team also includes Dr. Giovanni Da San Martino, Dr. Firoj Alam, Shaden Shaar, and Yifan Zhang, as well as several collaborators from the Arabic Language Technologies group, including Fahim Dalvi, Dr. Nadir Durrani, Hamdy Mubarak, Dr. Ahmed Abdelali, Dr. Hassan Sajjad, Dr. Kareem Darwish, as well as Alex Nikolov (intern from Sofia University). The team also features experts from the Social Computing group at QCRI including Dr. Ferda Ofli, Dr. Muhammad Imran, and Umair Qazi. 

The scientists developed a system for analyzing COVID-19 related social media messages in Arabic and English, modeling the perspectives of journalists, fact-checkers, social media platforms, policy-makers, and society as a whole. Using their system, they analyzed tweets about COVID-19, with a focus on vaccines, in English and Arabic originating from Qatar from two periods: February 2020 – August 2020 and November 2020 – January 2021. For Arabic tweets, data from the first period was analyzed, as there were not enough tweets from the second period.

They found that Arabic tweets contained a lot of false information and rumors, some discussed a possible cure, and a very small number were spreading panic. In contrast, English tweets (in both periods) were mostly factual, contained many jokes, and rarely rumors, and virtually no panic. 

The experts further analyzed the degree of propaganda in the tweets. They found that while the Arabic tweets were almost propaganda-free, this was not the case for the English tweets. For February-August 2020, about one-third of the tweets were propagandistic, and for November 2020-January 2021, propaganda went a bit down to a quarter of the tweets.

Next, the experts analyzed which propaganda techniques were used. They found that for Arabic, about half of the tweets expressed doubt, and one-fifth used loaded language. For English, (a) for February-August 2020, one third of the tweets used loaded language, and propaganda techniques such as exaggeration, fear, name-calling, doubt, and flag-waving each accounted for 10% plus of the tweets, and (b) for November 2020-January 2021, half of the tweets used loaded language, while flag- waving, name- calling, and exaggeration each accounted for 10% plus of the tweets.

Finally, they looked into framing, and found that in Arabic tweets, health-and-safety was the dominant perspective, with the economy coming second. In contrast, in English tweets, the economic perspective came first, while health-and-safety was secondary for both periods.

Read the full case study titled, ‘COVID-19 and Vaccines: A Twitter Study About Qatar’ here.