Falsehoods, which can range from genuine confusion and mistakes (misinformation) to deliberate lies (disinformation), often accompany threats like the COVID-19 pandemic – in what the World Health Organization (WHO) labelled an “infodemic”.

This makes it hard for decision-makers, healthcare workers, and the general public to distill trustworthy and reliable guidance. These challenges breed anxiety and confusion, and can even affect decision-making.

As a member of the Africa Infodemic Response Alliance (AIRA), UN Global Pulse partnered with WHO to provide insights on online misinformation and disinformation that put people at risk, in the context of COVID-19 and other public health concerns. We leveraged our groundbreaking radio mining project to tap into public radio conversations, as well as an out-of-the-box social listening tool to track social media discussions.

Our impact

Reports tracking confusion, misinformation, attitudes and concerns about the virus and vaccinations helped to inform the decision-making and outreach efforts of the WHO and other AIRA members. WHO was able to refine messages and better target information campaigns to fight the pandemic.

Our lessons

  • Off-the-shelf tools can be powerful for analysing social media and radio conversations, but they also have limitations and need to be adapted for the context in which they are deployed. For example, we trained a custom classifier to identify positive and negative discussions of the World Health Organization, since our tool’s built–in sentiment analysis was not sufficient. We also had to come up with creative strategies for analysing radio transcripts, since some of the automatic speech recognition tools did not have the word “COVID” in their vocabulary, or frequently mistranscribed it with other words (like “clover”, “covet”, etc.).
  • Like social media data, radio data has biases and shortcomings. However, we were able to show that radio conversations reflected emerging health concerns (like the rise of the omicron variant), as well as distinctions between countries (for example, differences in discussions of vaccines and blood clots depending on which vaccine a country had adopted).