Wikipedia is the perfect Covid-19 datasource, says IBM research director for healthcare informatics

Scientists are turning to an unconventional source to track coronavirus (Covid-19) cases in preparation for a new pandemic: Wikipedia. The open-source platform, which is open to volunteer-based editing and contributions, has proven to be an effective tool in monitoring and recording social and governmental cases around the world.

“With the pandemic, it was clear that regions and governments needed to intervene, but it wasn’t clear what would be the ultimate intervention,” said Dr. Michal Rosen-Zvi, research director of IBM Research speaking with CTech. “We realized that we could have learned from what different governments decided to do, but to be able to learn from that you need to have structured data. What we’ve done is collect data from 261 different countries and territories and we created a structure for this data.”  

Dr. Michal Rosen-Zvi Credit: Miri Davidovich

Rosen-Zvi discovered that thousands of people were crowdsourcing global information about Covid-19 in the most unlikely, but obvious, location: Wikipedia. Using each country’s dedicated page to the pandemic, IBM could then collect all the necessary data to help prepare for future pandemics, such as timelines and methods of intervention for things like transportation and public services.

Dr. Rosen-Zvi specializes in machine learning, medical image analysis, and cognitive computing. As an expert in Artificial Intelligence for healthcare and machine learning, medical imaging analysis, and bioinformatics, she knows how tech can save us if we’re ever faced with another pandemic.

Her appearance at the conference comes off the back of the publication of her latest paper, ‘The case of missed cancers: Applying AI as a Radiologist’s Safety Net’. In it, she argues against the idea that AI can replace radiologists entirely, but instead empower them with a tool to enhance their capabilities and provide patients with better, more efficient healthcare.