UVA Engineering earns a role in a National Science Foundation-funded center leveraging data science, big data analytics and visual analytics for industry, government
greenwood@cstone.netEvery day we produce 2.5 quintillion bytes of data, but the paradox of the Age of Big Data is that data by itself is meaningless. It is only when data is put to use for better decision-making that it becomes valuable.
That’s the rationale behind the Center for Visual and Decision Informatics, a National Science Foundation-sponsored partnership that brings the expertise of data scientists and engineers from a number of universities to bear on pressing challenges facing industry and government. The University of Virginia joined the center in March.
“The center’s activities are wide-ranging and designed for maximum impact,” said Peter Beling, associate professor of systems and information engineering and the center’s site director at UVA. Its mission is to develop next-generation visual and decision support tools and techniques that enable decision makers to significantly improve the way their organizations’ information is interpreted and exploited.
The center already has produced results. For instance, the center’s breakthrough techniques for forecasting the spread of influenza and predicting crime and disease hotspots will enable hospitals and police departments to more effectively allocate their resources.
The center is designed expressly to promote this sort of innovation. As part of its Industry/University Cooperative Research Center program, the National Science Foundation sets the stage by providing the center’s operating and overhead funds. Membership dues from industry members support a research agenda set jointly with their university partners. The NSF’s decision to award the Center for Visual and Decision Informatics its second five-year grant in February allowed it to expand.
“As a result of the NSF’s vote of confidence, the center now is among the largest of its kind in the United States,” Beling said. “It is the only one that focuses on data science, big data analytics and visual analytics.” Not surprisingly, the center has attracted an outstanding group of global companies that include Intel, Johnson & Johnson, Capital One, Thomson Reuters and Northrop Grumman. The University of Louisiana at Lafayette and Drexel University founded the center in 2012, and UVA and Stony Brook University joined when the grant was renewed.
The center also has a Finnish component. Tampere University of Technology became a university partner in 2014, Nokia is an industry member, and Tekes, the Finnish Funding Agency for Innovation, is a sponsor.
Real-World Challenges Drive UVA Contributions
“The center’s process matches the research paradigm that we follow in our department,” UVA Systems & Information Engineering, Beling said. “We start with an ambitious, real-world problem and develop a computational approach to the data that helps us address it.” UVA already has begun work in several areas where faculty members believe they can contribute to the center. Among them are credit finance and systemic risk. UVA researchers are developing models that will reinforce Capital One’s systems for detecting credit card fraud. With Northrop Grumman Technology Services, Professor William Scherer is developing methods of assessing systemic risk that can help market regulators foresee when a financial market is becoming unstable so that they can step in.
Another domain is one that Beling calls computational law. At the behest of Thomson Reuters, center researchers at UVA are developing techniques to provide decision-making support to help people interact with complex sociotechnical systems like the law. For instance, such a system could help people who couldn’t afford a lawyer navigate the personal, social and financial implications of filing for bankruptcy. UVA Associate Professor of Law Michael Livermore and Assistant Professor of Systems and Information Engineering Matthew Gerber are collaborating on the project.
With UVA Engineering’s new Link Lab attracting a score of researchers for collaboration on cyber-physical systems, that field is a natural contribution for the center. Beling is working with his colleague Assistant Professor Nicola Bezzo to develop sophisticated procedures for predicting the behaviors of unmanned aerial vehicles. “You can imagine a drone crossing the Potomac and heading toward downtown D.C.,” Beling says. “We want to create a system that can quickly help us gather information about its intent.”
Manufacturing is a longstanding interest for Beling. He is working with another center industry member, the Korean company UANGEL, which creates Internet of Things sensor platforms for factories. Currently, the information generated by these systems is used primarily for predictive maintenance. Beling is developing methods to move factory-floor information up the organization so it can be used for higher-level strategic decisions. “We know a lot about what is happening at the machine level, but this information is not being used by corporate executives,” he said.
An Opportunity to Tackle a Wide Range of Data Science Challenges
Other projects center researchers are tackling are less domain specific. UVA Psychology Professor Steven Boker and the center’s UVA Site Managing Director Stephen Adams, a research scientist in the Systems & Information Engineering Department, are researching methods for learning from data in ways that both preserve the privacy of data sources and the control the risks associated with sources that become untrustworthy because of cyber-attack or sensor failure. Other center projects by Systems and Information Engineering Professor Donald Brown and Computer Science Assistant Professor Hongning Wang are being developed.
Beling sees UVA’s entry into the center as a way to expand opportunities for faculty and students to develop innovative solutions to difficult real-world problems and to work side-by-side with some of the best companies in the world. “We want to break new ground,” he said. “We want to extend insight from data into regions where it hasn’t been possible before.”