AI project delivers key pandemic data to policymakers in Africa
At the centre of the predictive modeling and forecasting project is the creation of AI and data-powered dashboards, which display locally relevant information for policymakers as well as the public. These models serve as the official dashboards in Botswana, Nigeria, Eswatini, Mozambique, Namibia, Zimbabwe and South Africa, where more than 20,000 people view them every day. They provide locally nuanced analyses including:
- monitoring of COVID-19 transmission
- prediction of resurgences identification of emergent hotspots and outbreaks
- identification of individuals at higher risk of infection
- development of strategic, highly targeted and staged delivery plans of vaccines to priority areas
- monitoring to ensure public health interventions are equitable and effective for everyone.
Project researchers use the findings to inform COVID-19 policies and vaccine roll-out strategies, often working directly with policymakers and country taskforces. They also use findings to support communication strategies that address misinformation about COVID-19 vaccines, prevention and treatments, to improve people’s knowledge about what works and what might be dangerous.
"At the onset of COVID-19, there was a massive amount of information in Africa, and a lot of it was either competing or contradictory,” said Jude Kong, an assistant professor at York University, Canada, and the director of the Africa-Canada Artificial Intelligence and Data Innovation Consortium (ACADIC). “We felt compelled to help deliver the scientific rationale for public health interventions in a manner the layperson could understand.”
Kong said that the project’s success so far was due to several factors: "Stakeholder engagement, artificial intelligence and collaboration are instrumental. ‘Leapfrogging’ of novel ideas across countries and circumstances as well as consultations with ’traditional rulers’, policymakers and civil society organizations helped fill the data gaps and informed our models." Kong stated. “All this meant that we were able to have access to locally relevant data to address the rumours that were making rounds in African social media groups about COVID-19, COVID-19 treatment and vaccines."
An example of how the dashboard project has been working in South Africa can be found here.
The project is run by the ACADIC. Based out of York University, ACADIC brings together an interdisciplinary team of data scientists, epidemiologists, physicists, mathematicians and software engineers, as well as experts in AI, disaster and emergency management, clinical public health, citizen science and community engagement.
The project is part of the Global South AI4COVID Program, which supports multidisciplinary research focused on evidence-based AI and data science approaches to aid COVID-19 response and recovery in low- and middle-income countries. The program is funded by IDRC and the Swedish International Development Cooperation Agency (Sida), with support from Pulse Lab Jakarta of the United Nations Global Pulse network, Gender at Work, and Ladysmith.