Assistant Professor Adam Poliak and Joon Luther ’25 collaborated on a system for identifying discussions of shingles vaccine side effects in social media posts. Their work combined several machine-learning models to classify more than 3,000 labeled Reddit posts, exploring how online conversations can offer timely public-health insights.
Research co-authored by Assistant Professor Adam Poliak examines the public-health consequences of rapidly expanding online sports betting. Using Google search trends, the team found a sharp rise in searches associated with gambling addiction following the 2018 legalization of sports betting in many states.
Grace Proebsting, a Haverford College student who graduated in 2025, and Assistant Professor Adam Poliak published “Biases in Large Language Model-Elicited Text: A Case Study in Natural Language Inference” at COLING 2025.
Jac McCarty successfully defended her interdisciplinary thesis, “Russian Language in Modern Natural Language Processing and Natural Language Inference: The Necessity of Performance Evaluation Datasets Focusing on Negation,” on Thursday, May 9, 2024.
Jac McCarty ’24, a Russian and computer science double major whose undergraduate research examined natural language models in Russian, was among the Bryn Mawr students selected for a Boren Scholarship.
In an interview with Barnard Computer Science, Adam Poliak discusses his research on the capabilities and limitations of natural language processing systems, including the ways biases in models and datasets can affect real-world technologies.