Undergraduates Fight Echo Chambers with Evolutionary Computing

A team of Manning CICS undergraduate student researchers, advised by doctoral student Cooper Sigrist and including team lead and co-lead author Ryan Boldi, co-lead author Aadam Lokhandwala, Alexander Lavrenenko, Yuval Shechter, and Edward Annatone, won the 2022 ProjectX Human-Computer Interaction Prize as part of a five-month international machine learning research competition hosted at the University of Toronto.

Their winning paper, “Recommendation Diversity Worth Caring About,” focuses on recommendation systems—algorithms that control the content users are exposed to while shopping, researching, or browsing social media. Recommendation systems are designed to help users find what they’re looking for faster—but the more they learn about users, the more they narrow the scope of suggested content, risking the creation of echo chambers and the deepening of political and social divides. “Our team was concerned about how dangerous poorly optimized recommender systems can be,” explains Boldi. “With social media at the center of controversial and extreme political polarization, we thought recommender systems would be an especially important and relevant focus for our research.” 

To help solve the problem, the team introduced a “serendipity” metric to specifically target echo chambers and homophily in recommendation systems, an augmentation for recommendation systems to use the new metric to preserve the display of diverse content, and the use of lexicase selection—a strategy for evolutionary computation known for its ability to spontaneously promote diversity without relying on explicit diversity clues. Boldi brought the idea of using lexicase selection to the team from his previous research work with Lee Spector, Amherst College professor of computer science and CICS adjunct, with whom he has published eight papers on the subject and presented at conferences in Portugal and Japan. 

These projects directly feed Boldi’s current research goal—to study the emergence of human intelligence from the perspective of evolving and learning autonomous systems—which he plans to pursue as a computer science doctoral student starting in 2025. “Research has been vital to my development as an undergraduate student,” he says. “Not only has it been fulfilling, but it has taught me invaluable communication, teamwork, problem solving, and programming skills.”

This story originally appeared in the Winter 2024 issue of Significant Bits magazine.