UD’s Data Science Institute Harnesses Data as a Tool for Good

With the world awash in information, the Data Science Institute (DSI) based at the University of Delaware serves in the vanguard of research into the many ways data can affect and assist society.

The institute “is a networking nucleus that brings faculty and students together,” drawing from all the university colleges, says Cathy Wu, the institute’s founding director.

“This allows us to accomplish things we wouldn’t be able to do alone within our own disciplines, because we need collaboration with researchers in other fields,” adds Wu, who is the Unidel Edward G. Jefferson Chair in Engineering and Computer Science and a professor of computer and information sciences.

The DSI, with about 150 faculty members, also offers symposiums, programs and outreach that are helping to position Delaware as a leader in the data-science world. Its affiliations include projects aligned with the university’s FinTech Innovation Hub that allow access to the Institute’s resources by private and public organizations.

“We also affiliate with other educational partners like Delaware State University and Delaware Technical Community College, and have aliaison with industry partners,” adds Wu, a pioneer in international data-bases involving research in health and other fields; she has also served as an advisor to the National Institutes of Health.

“In terms of the broader community, we’ve initiated a new collaboration with the U.S. Department of Veterans Affairs,” Wu says. “We’re working on how we can bring the knowledge we’re collecting through data science, including the social determinants of health, to help with critical issues the VA wants to address, such as suicide prevention.”

Dealing With the AI Revolution
Two key topics that many data-science experts grapple with are cybersecurity and artificial intelligence. “Cybersecurity is definitely a big issue,” says Wu. “And in collaboration with the university Center for Cybersecurity, Assurance and Privacy, the DSI is exploring it from several areas, including the engineering and financial services side.”

When it comes to AI, she notes: “Machine learning and AI have been in existence for decades, but with big data and high-performance computing, the acceleration in AI is unprecedented.”

As a result, she says, the DSI is ramping up plans to expand AI research and education, in close collaboration with the university’s Artificial Intelligence Center of Excellence.

Benjamin Bagozzi, assistant director of the university’s Masters in Data Science Program, serves as an associate director of the DSI.“

Masters in Data Science Program Assistant Director Benjamin Bagozzi | PHOTO CREDITS: BENJAMIN BAGOZZI IMAGE COURTESY OF UNIVERSITY OF DELAWARE; SATALITE IMAGE OF DELAWARE BY NASA TAKEN BY MODI

One great feature of the Institute is that there’s a group within it that’s passionate about the ethics of AI and data science,” he says. He points out that the DSI holds symposiums and panels on the ethical implications of AI, and students in the Master of Data Science Program are required to take a course on ethics in data science and AI. “The power of these tools is immense,” Bagozzi says, “and we all want to be careful in how they’re used.”

Bagozzi, who also serves as an associate professor in the university’s Department of Political Science & International Relations, was originally trained in political science. “Nowadays, one of my key threads is in studying written text as data, in terms of social processes and politics.”

He explains: “I’ve analyzed various types of speeches, often at the international level, such as at the UN. The purpose is to try to capture how countries might be aligning with or against one another when it comes to opeating, or negotiating over issues like climate change.” He also researches and analyze show individual citizens and their governments interact with each other through the use of written text.

Unraveling the Complicated System of the Brain
Austin J. Brockmeier is a resident faculty member of the DSI; he is also an assistant professor in the Department of Electrical & Computer Engineering, and the Department of Computer &Information Sciences.

Austin J. Brockmeier
Assistant Professor,
University of Delaware

Brockmeier does in-depth data science research about the brain, working in collaboration with biomedical researchers and neuroscientists.

“The brain is obviously a complicate system,” Brockmeier says, “and I do analysis forclinical and neuroscientific outcomes.The goal is to find what changes happen inthe brain based on images you get from anMRI scanner.” This type of specialized datacan help predict possible future medicaloutcomes, he adds: “For example, things likeearly indicators of cognitive decline or otherdisorders like schizophrenia.”

He continues: “If research finds patterns that are predictive of future events, it might be used to adjust behavior in a positive way. If the risk factors are modifiable, for instance, a person’s doctor might tell them something like, ‘Hey, you have a very high risk for this, so maybe you can exercise more, or eat better to help lower the risk.’”

One project that Brockmeieris participating in deals with data on mild traumatic brain injury. “We’re looking to see how it might have an effect later in life and collaborating with experts in concussion. Overall, we’re analyzing the possibility that any kind of traumatic brain injury does affect the structure of the brain; that it’s a bruise and doesn’t always heal, so we’re looking at the long-term outcomes of that.”

Using Data to Benefit Agriculture
Dr. Pinki Mondalis also a resident faculty member of the DSI, and an assistant professor in the Department of Geography & Spatial Sciences. In addition, she serves as director of the Environmental Science undergraduate program.

Dr. Pinki Mondal
Assistant Professor,
University of Delaware

“Most of the data I use comes from space satellites,” says Dr. Mondal. “NASA sends up satellites that obtain images of the earth and collect information for use by researchers all over the world. So there’s a huge amount of data, starting from the 1970s.”

Dr. Mondal’s background is in environmental geography. “When people think of geography, they usually think of maps. But as environmental geographers, we deal with hundreds of gigabytes of computer data. To handle all that information, we need data science training that will help us, in turn, to ‘train’ the computers to tell us what’s going on with the satellite images.”

One of Dr. Mondal’s projects concentrates on Delaware’s agriculture. “We’re analyzing images taken by satellites, as well as from drones and aircraft, to look at how saltwater is flowing in from the ocean and affecting Delaware farmland. We combine all the information and ‘train’ the computer using simplified AI models, because it’s a huge amount of data and we need more than just human power. And we need to do it fast, because we need to tell our climate communities and our farmers what’s going on.”

Dr. Mondal believes in the importance of openly sharing data, “so anyone can go to our research website and download the data on saltwater intrusion in Delaware; we’ve also created an app for it.”

She adds: “These are urgent conditions that affect Delaware’s economy, because many farms here grow corn and soybeans, and salt-water can really impact the state’s agricultural production.”