Skip Navigation

Scout Archives

Home Projects Publications Archives About Sign Up or Log In


The breadth of research surrounding COVID-19 can be overwhelming, but SciFact helps researchers sift through studies to find relevant existing work. Before using the resource, readers should be aware of a few disclaimers. First, "the tool is not a substitute for the care provided by licensed healthcare practitioners." Second, the system is designed to assess whether scientific research "support[s] or refute[s] a scientific claim," but does not "assess the credibility of the scientific paper itself." Thus, readers should be sure to further investigate each paper they find. The SciFact system itself is very intuitive. Users type a claim in the search bar (for example, "masks reduce COVID-19 transmission,") and the engine returns a list of scientific papers that support or refute the claim. Based on the example search, users would find summaries of seven research papers that support the claim. How does the system generate these results? After it receives an "input claim," the system combs through COVID-19 research for relevant abstracts, identifies relevant evidence within that abstract, and then assesses whether the abstract supports or refutes the claim. SciFact is a project of the Allen Institute for Artificial Intelligence, a non-profit research organization.
Archived Scout Publication URL
Scout Publication
GEM Subject
Date of Scout Publication
November 20th, 2020
Date Of Record Creation
October 26th, 2020 at 6:24pm
Date Of Record Release
November 16th, 2020 at 12:18pm
Resource URL Clicks
Add Comment


(no comments available yet)