Leveraging Social Media for COVID-19 Response: Insights from a Data Competition

The COVID-19 pandemic accelerated the adoption of digital platforms across various sectors, notably in education and healthcare, with remote learning and social media emerging as pivotal tools for communication and crisis management. Social networks played a crucial role in disseminating critical information, combating misinformation, and fostering community engagement. Recent research underscores the significance of social media in shaping public behavior towards adopting protective measures against COVID-19, yet quantifying its precise impact remains challenging due to the complexity of social relationships and diverse information sources. Multimodal data generated by social media platforms presents opportunities for more insightful Machine Learning (ML) models, but also poses technical challenges in data integration and interpretation. Leveraging crowdsourcing, we organized a data science competition aimed at forecasting COVID-19 positivity rates and identifying factors influencing its spread using infection and social media data. The competition facilitated collaborative problem-solving and provided actionable insights for public health communication and policy-making. This study outlines the competition structure, methodologies employed by participants, key findings, and implications for future pandemics and public health crises.