Two University of Dayton computer scientists were awarded National Science Foundation grants totaling nearly $350,000 through a program that supports research among early-career faculty or research scientists.
The funding also will provide cutting-edge research opportunities for graduate and undergraduate students.
Ngoc Nguyen, assistant professor of computer science, was awarded $174,592 to create an AI-powered educational game platform to teach collaborative decision-making with adaptive feedback using cognitive and large language models.
Tianming Zhao, assistant professor of computer science, was awarded $174,991 to study the feasibility of impersonation attacks on biometric authentication systems in mobile and wearable devices and develop practical methods for defense.
The grants were awarded through the NSF’s Computer and Information Science and Engineering Research Initiation Initiative (CRII).
“The NSF CRII grant is the initial spark that ignites an independent research career,” said Michael Poor, associate professor and UD Department of Computer Science chair.
“It provides essential resources, such as funding for their very first Ph.D. student, enabling them to move beyond preliminary ideas and establish a robust, independent research agenda early in their tenure-track journey,” Poor said. “This grant fundamentally changes the trajectory of their scholarship.”
Poor said junior faculty face immense pressure to balance teaching, service and research. The CRII award directly addresses this workload challenge by providing dedicated student support, allowing qualified graduate students to assist with core research tasks. This enables them to manage their teaching duties while still generating the rigorous, high-quality preliminary data needed for successful tenure review.
Nguyen joined the UD faculty in 2023 after serving as a postdoctoral fellow at the Dynamic Decision-Making Lab at Carnegie Mellon University. Her research focuses on understanding how people make decisions amid uncertainty, and how to design artificial intelligence systems that effectively support and collaborate with humans in individual and group decision-making.
“Making decisions as an individual is tough, but in a group setting it is even more complicated because of the dynamics of people — different personalities, preferences and factors come into play,” Nguyen said.
Her AI-powered game platform will allow students to practice making collective decisions in a group setting to help them see the value of teamwork. It also facilitates teachers, who can design or upload decision-making scenarios, from complex business negotiations to deciding where to go for dinner.
Her platform will use cognitive modeling — enhanced by large language models, which are the backbone of ChatGPT and generative AI — to infer each player’s priority or intention based on the discussion and provide real-time feedback.
The platform also will use strategy simulators with reinforcement learning agents to look at “what if” scenario analytics to see how various options might play out.
“Once this platform is built as a prototype, I’m going to use it in my class, where we can form a group of students to test it and then collect the data to see the dynamics of human behavior when we’re making decisions as a group,” she said.
Nguyen’s funding will support one graduate and one undergraduate student to work with her on the project.
Zhao joined the UD faculty in 2022 after completing his doctoral degree in computer and information science at Temple University. His project uses heart rate-monitoring photoplethysmography (PPG) sensors on smart watches for continuous user authentication, instead of a password or fingerprint pattern.
“The idea is that different people have different cardiac patterns that can be captured by this PPG sensor,” Zhao said. “I use that for authentication.”
To create a robust cyber security system, Zhao not only considers device authentication, but also a potential attacker’s perspective. For example, he developed a practical attack procedure called remote cardiac sensing, which analyzes video of a person’s face taken from a distance at the pixel level.
“The pixel color changes in a pattern associated with your heartbeat,” he said.
His research will explore how to transform this remote-captured cardiac pattern into a signal suitable for smart watches and other wearable devices using machine learning and artificial intelligence. He also will explore how to compromise the authentication system to inject the cardiac pattern data into a smart watch and assume the identity of the person who is supposed to be wearing the watch.
As required by the NSF, Zhao also will propose a corresponding defense method to detect and repel such attacks. He has hired an undergraduate student to work on the project and plans to recruit others to assist with its various components.
Poor said the NSF CRII grants are the “vital first step on the federal funding ladder,” serving as a proving ground for early-career faculty to establish a track record of excellent research, effective project management and successful communication with a federal agency. This lays a critical foundation that makes them highly competitive for the prestigious, 5-year NSF Faculty Early Career Development (CAREER) award in the future.
“Securing long-term, multi-million-dollar funding like the CAREER would be monumental for the faculty member and a huge feather in the department's and the University's cap,” Poor said.
Top photo: Tianming Zhao
Middle photo: Ngoc Nguyen