Doctoral Student Embodies Interdisciplinary Approach to Research

KENNESAW, Ga. | Aug 24, 2020

Much like the winding path that led her to Kennesaw State, Liyuan Liu’s approach to research has been anything but linear.

Since starting in the University’s Ph.D. in Analytics and Data Science program in 2017, she has published 18 peer-reviewed research papers alongside faculty representing three Kennesaw State departments: computer science, information technology and psychology.

Though on paper it may appear scattered, Liu said the interdisciplinary approach has provided her a more holistic understanding of how to apply research techniques as she explores how data-driven approaches can help inform companies in making business decisions and how to encourage users to take part in the data-sharing process.

“No matter the subject, we can always use data analytics to solve real-world problems,” she said. “Because of this, I found it very easy to integrate with professors across multiple departments and learn about how these approaches can be applied to their respective fields.”

Like many who study in KSU’s Analytics and Data Science Institute, Liu came to the University with a diverse educational background. After earning an undergraduate degree in business administration with a concentration in human resource management in China, she moved stateside to pursue graduate degrees in financial economics and analytics at the University of Detroit Mercy and Georgia State University, respectively.

After discovering a passion for data analytics, Liu said she began to search for doctoral opportunities before stumbling across a webpage for KSU’s program, the first of its kind in the country. After speaking with Jennifer Priestley, executive director of the Analytics and Data Science Institute, Liu said she was enthralled by the theoretical approach to data analytics, the institute’s deep connection to industry partners and the freedom to conduct independent research.

Through the institute, Liu has conducted work in the Equifax and Novelis research labs. With Equifax, she helped develop a credit model utilizing different advanced machine learning techniques. In the Novelis sponsored lab, she helped identify actionable items that allowed the company to make managerial decisions using a data-driven approach.

“Since her arrival, Liyuan has truly been the embodiment of how interdisciplinary the Ph.D. in Analytics and Data Science program can be,” Priestley said. “She approaches every new challenge with an open mind and constantly innovates new ways to approach everyday problems.”

In what would become her favorite research project, Liu used an artificial intelligence algorithm to detect sarcasm in text messages. While it is easy to pick up sarcastic comments when spoken in the presence of another through inflection or physical cues, she said machines often miss context clues in text, such as punctuation. Her approach included these clues to better detect sarcasm.

Originally an oral presentation delivered as part of her coursework, Priestley later encouraged Liu to continue researching the topic in the form of a paper. Her work would go on to earn best student paper at the 2019 IEEE International Conference on Cognitive Machine Intelligence. She previously won first place in the graduate capstone division of C-Day, a computing showcase held each semester by the College of Computing and Software Engineering.

Upon earning her doctoral degree, Liu said she would like to pursue a career in academia.

“I have always enjoyed teaching,” she said. “By remaining in academia, I would be able to chase one passion while continuing my research. It’s a win-win.”

– Travis Highfield

Photos by David Caselli

Doctoral Student Embodies Interdisciplinary Approach to Research

Related Stories