Fourth-year Computer Science student Tyler McDonald was packing his bags for an intentional conference in Abu Dhabi, United Arab Emirates, when he got some exciting news.

The research paper he co-wrote with Angel Loredo Lopez, an undergraduate student from Mexico who had been at Brock on a scholarship program, had won the Best Dataset Paper prize from the 2025 International Conference on Computational Linguistics (COLING).

Assistant Professor of Computer Science Ali Emami says these international conferences accept only about 20 per cent of the thousands of submissions, which typically originate from leading institutions worldwide.

“As undergraduates, we developed research that received recognition on an international stage,” says McDonald. “This achievement wouldn’t have been possible without the exceptional support from the Department of Computer Science.”

Emami and his colleagues actively seek to give graduate and undergraduate students opportunities to engage in cutting-edge research.

“Throughout the process, I stay attuned to their emotions and challenges, and when the work succeeds, I help them recognize it was really their capability driving it forward all along,” he says. “The most rewarding part is watching students gain confidence with each project, gradually taking more ownership and initiative as they discover their own potential.”

It was through a collaborative research opportunity that then-undergraduate student Robert Morabito (BSc ’24) connected with McDonald and fourth-year Computer Science student Sangmitra Madhusudan.

The trio earned the Social Impact Paper Award at the 2024 Conference on Empirical Methods in Natural Language Processing. Among the more than 6,000 submissions, only Brock’s paper and two others received this distinction.

The team’s research, “STOP! Benchmarking Large Language Models with Sensitivity Testing on Offensive Progressions,” developed methodology for evaluating how large language models such as GPT-4, Mixtral and Llama 3 detect biases underlying racist and offensive content.

Their findings revealed that models detect biases inconsistently and demonstrated how their framework can help create fairer language models by identifying and mitigating these biases.

Morabito initially started the research on his own but faced significant challenges. Rather than giving up, he sought out Madhusudan for her expertise in fine-tuning language models and McDonald for his writing proficiency.

“Dr. Emami encouraged me to provide feedback on Sangmitra’s projects, while Sangmitra shared her expertise in fine-tuning methodologies,” says Morabito. “There’s a genuine culture of mentorship, with professors and undergraduates learning from each other.”

Madhusudan says Emami’s passion is evident in the feedback he provides.

“He consistently encourages us when we encounter obstacles and supports our learning through trial and error,” she says.

Computer Sciences Department Chair Beatrice Ombuki-Berman says the award-winning achievements reflect the department’s commitment to involving students at all levels in meaningful research.

“By pairing faculty expertise with student creativity, we’re building a collaborative community where students don’t just study computer science—they actively advance it,” she says.