About Me

CV (Current as of 01/2024): [Click to Download]
Google Scholar: Link
DBLP: Link

Bio

Hi, my name is Joel. I’m a Computer Science PhD student at the University of Waterloo, supervised by Dr. Lukasz Golab. My research is focused on developing novel explainable artificial intelligence (XAI) techniques for new applications of deep learning (e.g., LLMs and ranking models), leveraging a variety of analytical methods (e.g., natural language processing, rule mining).

Recently at ICDE 2023, we presented our first efforts in adapting counterfactual explanations for information retrieval (IR). In this work, we propose the first counterfactual explanation formulations for document ranking models, presented in an interactive explainability tool named CREDENCE. For more info, check out our preprint here. We are currently generalizing our focus to large language models (LLMs), studying how they learn from in-context knowledge sources during retrieval-augmented generation (RAG).

I received my Bachelor of Computer Science (Honours) from the University of Windsor, graduating in May 2019 with Great Distinction. Under Dr. Alioune Ngom, I spent several years researching and implementing machine learning models for breast cancer treatment and diagnosis. I also studied natural language processing (NLP), under Dr. Robin Gras, in which we adapted the (then brand-new) BERT language model for sentiment analysis on social media posts.

Outside of school, I work as an Engineering Manager for Movyl Technologies, where I am responsible for a wide variety of data engineering and data science efforts. Previously, I worked as a Software Engineer for the Federal Government of Canada, in Ottawa ON.