Foundation selects Matt Turk as Moore Investigator in Data-Driven Discovery

By Dave Evensen, College of Liberal Arts and Sciences

A $1.5 million grant will help develop a tool to assist scientists in interpreting research data and make discoveries faster, says an astronomy professor and research scientist at the National Center for Supercomputing Applications (NCSA).

Matthew Turk has been named a recipient of a Moore Investigator Award in Data-Driven Discovery from the Gordon and Betty Moore Foundation. These awards are part of the foundation’s Data-Driven Discovery Initiative, which is committed to enabling new types of scientific breakthroughs by supporting interdisciplinary, data-driven researchers.

Turk plans to use the five-year award to further develop the “yt project,” which he founded to help his own study of early stars.

“One of the things that can be very difficult with data particularly from computer simulations is that it is often written in form that is meaningful to the simulation code or platform, and not to scientists,” Turk says. “So what we’ve attempted to do with this tool is to foster the ability to ask questions that are physically meaningful rather than just computationally meaningful.”

Turk, who arrived at the U of I last June, developed yt as a graduate student as he struggled to analyze raw research data. Yt is basically a tool with powerful, Python programming language for “ingesting, indexing, and processing data,” he explains, used on equipment ranging from mobile devices to supercomputers. It’s creates a single environment to conduct data analysis, visualization of data, and even planetarium shows.

“Many areas of science are currently data-rich, but discovery-poor,” says Vicki Chandler, chief program officer for science at the Moore Foundation. “The Moore Investigator Awards in Data-Driven Discovery aim to reverse that trend by enabling researchers to harness the unprecedented diversity of scientific data now available and answer new kinds of questions.”

Turk is quick to point out that while he launched yt himself, other research scientists quickly collaborated with him and helped develop the tool further. With the award, however, Turk plans to hire staff to further develop the project so that the tool can be used by everyone from students to faculty in a wide range of research, from studying stars to analyzing the flow of water. Part of the award will be used to conduct annual workshops for using the tool. Brian Fields, chair of the astronomy department, says that Turk’s work comes during a “golden age” for astrophysics and cosmology, as improving telescopes provide scientists with a flood of hard data about the universe. The flood of data has also presented some challenges.

“Increasing computer power has allowed astrophysicists like Matt to build large, sophisticated simulations of the cosmos, and it is increasingly challenging to compare the predictions from these simulations with observational data,” Fields says. “Those challenges face everyone. Matt saw them in his work, and yt will help address those challenges in a user-centered way, to make it easier for users to ask science questions.”

Ed Seidel, director of NCSA, says Turk is a promising, central part of NCSA’s push toward data-intensive science and engineering. His work, Seidel adds, is applicable across many disciplines.

“He’s unusually broad,” Seidel says. “I’d say he’s a new breed of scientists that works as easily in astrophysics as he does in data-intensive techniques and tools, and I think you see that more and more when you have interdisciplinary approaches that are touching every field. Particularly, the younger generation of scientist has this ability to go between disciplines more easily. He’s a perfect exemplar. That’s why we were glad we could attract him to NCSA, and that’s why I think the Moore Foundation has recognized him.”

For more information about the Moore Investigators in Data-Driven Discovery, please visit http://www.moore.org/programs/science/data-driven-discovery/investigators.

Disclaimer: Due to changes in website systems, we've adjusted archived content to fit the present-day site and the articles will not appear in their original published format. Formatting, header information, photographs and other illustrations are not available in archived articles.