Supervisors: Graeme Hirst, Frank Rudzicz
Affiliate of the Vector Institute for Artificial Intelligence
Co-founder and Technical Developer at Mutuo Health Solutions
My current research is focused on health applications of natural language processing (NLP), including automated cause of death classification for verbal autopsy reports and temporal ordering of events in medical narratives.
I previously worked at Safaba Translation Solutions and Amazon's Alexa team.
I received my Master's degree in Language Technologies from Carnegie Mellon University, and my Bachelor's degree in Computer Science from Johns Hopkins University.
Verbal autopsies are an alternative to a medical autopsy in cases where deaths occur away from hospitals and no physical autopsy is performed. Keeping track of the causes of these deaths helps countries like India make decisions about public health policy and funding. Our current work focuses on automatically determining the cause of death from the narrative (text) portion of a verbal autopsy record using NLP and machine learning.
The Million Death Study
Automatically determining cause of death from verbal autopsy narratives (2019)
Multi-task learning for interpretable cause-of-death classification using key phrase prediction (2018)
We built an English to Egyptian Arabic Machine Translation system using the Moses toolkit and data from CMU Qatar.
Arabic NLP Workshop Paper (2014)
Preprocessing code: https://github.com/sjeblee/egymt
Project Members: Weston Feely, Alon Lavie, Houda Bouamor