SMU Computer May Help Speed Coronavirus Solution

The computer culls through thousands of medical journals, news articles and other sources in search of clues.

NBCUniversal, Inc.

Students and faculty at Southern Methodist University in Dallas are part of a global effort to find a coronavirus vaccine.

They’re using a giant computer on campus to cull through thousands of medical journals, news articles and other records, that no human could possibly read, in search of clues.

Jo Guldi is an SMU associate professor of history who specializes in mining text on social issues with artificial intelligence.

“It's a very niche, sort of cutting-edge methodology. There aren't that many of us in the world,” Guldi said.

She is a co-founder of the SMU Artificial Intelligence Lab, along with Frederick Chang, an SMU cyber-security professor and computer science expert.

They work with a mainframe computer they call the "Mane Frame" and branded like an SMU Mustang

The digital beast is housed in a campus building of its own at a location that was once a bread factory. The new building cooks data instead, in what Chang considers a fine example of the amazing advance in artificial intelligence over the past 10 years.

“So it's AI, modern AI, versus coronavirus,” he said.

The project also benefits from available students who have none of the typical campus life to distract them during coronavirus stay-at-home recommendations.

“They have all this energy and we're all anxious about what the coronavirus is going to do to the economy and to society,” Guldi said.

The professors said they found it easy to recruit student help for the project.

“We have the opportunity to work on something larger than ourselves. We're happy to do it,” Chang said.

The team has "Mane Frame" speed reading around 52,000 scientific papers in search of patterns that may help coronavirus researchers. Other institutions around the world are also involved.

“We're sort of augmenting the intelligence of a bunch of really smart infectious disease researchers, who just don't have time to read 52,000 papers,” Chang said.

The experts said the computer is not smart enough to actually develop a vaccine by itself, but it could shorten the time it takes humans to do it.

“So that's what I'm excited about,” Guldi said.

They hope to see early nuggets of genius from the artificial intelligence analysis along the way.

Contact Us