Misinformation in Africa and the rest of the Global South often takes the form of extreme speech that incites violence or spread racist, misogynous, xenophobic messages. Since the start of the COVID-19 pandemic, the proliferation of fake news and misinformation has been a constant battle for health officials and policy makers as they work to curb the spread of COVID-19. According to Statista, social media generates 42% of the traffic related to misinformation and fake news. The dynamic nature of platforms such as Twitter, the tweets based on their respective Facebook, and WhatsApp allows for misinformation to be spread quickly and widely. To address this, techniques such as Latent Dirichlet allocation (LDA) or BERT could be used to extrapolate topic modeling on messages for detecting fake news that could be viable, allowing governments and health organizations to identify topics they could focus on.
Future work involving the dataset we collected would be to categorize the tweets based on their respective representations.