Accelerated master’s student Charles Cutler explores the diverse applications of data science through cross-departmental collaboration
For Cutler, data science goes beyond the boundaries of computers and into fields where innovative research is changing lives.
Accelerated master’s student Charles Cutler has spent his time at the VCU College of Engineering investing in collaboration. From his senior capstone project working with VCU Health and the VCU School of Medicine, to working alongside chemical engineers and their complex patents, Cutler is bringing innovative data science solutions to difficult problems across many disciplines.
Cutler is in his final year of graduate studies and, after bringing approximately 40 credits from his high school dual-enrollment courses, will have both his bachelor’s and his master’s degree in computer science.
“After looking at the accelerated master’s program at VCU Engineering, I was amazed that in four years I could have a bachelor’s degree, master’s degree and publish a thesis,” Cutler said. “That’s an opportunity I didn’t see anywhere else and what initially drew me to this program.”
As an undergraduate, he began working in the Natural Language Processing (NLP) lab alongside Bridget McInnes, Ph.D., spurring his passion for data science and training machines to solve complicated problems. Additionally, research in the NLP lab inspired Cutler’s current thesis project focused on continual learning within Named Entity Recognition (NER).
Named Entity Recognition is the process by which a NLP model identifies and classifies named entities (such as people, locations, organizations, etc.) within unstructured text. Continual learning is the ability of these models to then adapt and learn as new data becomes available or a new task is assigned. The main challenge with continual learning is something called catastrophic forgetting, or when a model loses previously learned information as it takes in new knowledge or tasks.
Cutler’s thesis explores ways to mitigate catastrophic forgetting in new, less expensive and more effective ways. One solution implements a student-teacher framework in which an older NLP model trains a new model, guiding the newer model to adapt to new entities while retaining the old information.
In addition to this system, Cutler utilizes generative AI to create synthetic data, minimizing the typical privacy concerns that occur when working with large datasets. This leads to a more robust, safe and extensive training process that improves the models’ adaptability to evolving language.
Cutler is using this same method to help chemical engineers sort through complex patent material, presenting his findings at the recent graduate poster symposium.
“We want to apply it to chemical patents so we can extract these really difficult chemical entities from the often vague patents, getting them out and into the chemical engineers hands,” Cutler said. “Papers are being published faster than we can read them. We want to be able to take any text, maybe it’s scientific literature or chemical patents, and extract useful information from it.”
Last year, Cutler and his senior capstone team collaborated with the Department of Occupational Therapy in the VCU College of Health Professions. Their sponsor within the department, Virginia Chu, Ph.D., aided their research, creating a system to automatically detect early motor skill deficiencies in children. This project ultimately won both the Sternheimer award and overall first place at the 2023 Capstone Design Expo.
“It was one of the most amazing experiences of my life,” Cutler said. “I’d never worked on a project with that real of an impact before, and being able to work with VCU Health was just incredible.”
With the majority of his work surrounding NLP, Cutler emphasizes the importance of speaking out about the ethical use of AI, fostering conversations among researchers and field professionals about responsible data usage and ways to develop more transparent algorithms.
“I’m an optimist about the future,” Cutler said. “But it’s important to talk about these things now so we don’t regret it down the road. It’s something I’m passionate about and would love to continue researching beyond graduation.”
After graduation, Cutler wants to continue researching to make an impact on the world. His current aspiration is working with the National Institute of Health or the National Library of Medicine, solving meaningful problems in the field of health and medicine.
“I love learning about all the unique applications of data science,” Cutler said. “The idea that you can teach a machine to complete tasks or make predictions about real world problems is amazing. With all the data that is around us, we can always be learning something new and finding innovative ways to solve problems.”
The Department of Computer Science provides undergraduate and graduate students with the opportunity to perform real-world research as soon as they enroll. From designing algorithms to solving complex computing problems to working with cutting-edge AI technology, students gain understanding of many important computing topics. Browse videos and recent news from the Department of Computer Science to discover how the College of Engineering at Virginia Commonwealth University prepares the next generation of scientists and engineers for the challenges of the future.
Categories Computer Science, Graduate Student Stories, Student Stories