AI Fights the Pandemic (And How You Can Get Involved)
A new AI challenge is poised to help fast-track research to mitigate the impact of COVID-19.
- By Lindsay Stares
- March 17, 2020
The spread of the novel coronavirus is now officially a pandemic, causing widespread disruption to normal life around the world. While hospitals ramp up capacity and governments encourage or mandate quarantines, the data and technology space is mobilizing to help.
New AI Challenge Announced
One opportunity is the COVID-19 Open Research Dataset Challenge (CORD-19), presented on Kaggle by a coalition that includes the Allen Institute for Artificial Intelligence (AI2), the Chan Zuckerberg Initiative, Georgetown University’s Center for Security and Emerging Technology, Microsoft Research, the National Library of Medicine, and the White House.
The challenge calls on AI experts to create “text and data mining tools that can help the medical community develop answers to high priority scientific questions” about the coronavirus. The provided data includes over 29,000 scholarly articles about COVID-19 and related viruses. Exploring the literature with AI techniques could greatly accelerate researchers’ ability to understand the virus and find viable therapies.
Pulling this data together and making it ready for natural language processing (NLP) and machine learning has been the work of an extraordinary collaboration between public and private entities. According to coverage on ZDNet, “Microsoft used its web-scale literature curation tools to pull together global scientific efforts and results. The NLM provided access to literature content, while the Allen Institute transformed the content into machine-readable [form].” The data set will continue to be updated as new research is published.
The project’s approach grew out of the Allen Institute’s existing work using NLP and machine learning to analyze large amounts of scientific literature. You can hear some key figures provide more background and context on a recent GeekWire Health Tech Podcast.
What Is Already Being Done?
Of course, the CORD-19 project isn’t the only way to use data and AI to fight the coronavirus. DeepMind has already released preliminary findings about related protein structures predicted by deep learning. Understanding the protein structure of a virus can give scientists a starting point for developing vaccines or drug therapies.
VentureBeat recently reported on several AI and tech-driven initiatives , including robots wielding disinfectant mobilized in China, cameras with thermal sensors for fever detection, and deep learning models that might diagnose patients faster or predict outcomes.
Meanwhile, general COVID-19 data such as numbers of cases and demographic and regional statistics has become ubiquitous online, giving rise to hundreds of dashboards and data visualizations of varying quality. The impulse to understand the data has been so strong that Tableau released a set of guidelines for data visualization makers to help prevent those without a background in epidemiology from accidentally misleading the public.
An Opportunity to Contribute
Are you an AI specialist ready to jump in and contribute? Read the initial key questions, find out about the cash prizes, and access the data on Kaggle.
Lindsay Stares is a production editor at TDWI. You can contact her here.