Tonya Custis leads a team of research scientists performing applied research in language and search technologies.
Tonya earned a Ph.D. in Linguistics, an M.S. in Computer Science, and an M.A. in Linguistics, all from the University of Minnesota. She also has a B.A. in Music from the University of Connecticut. Her research interests include information retrieval, natural language processing, statistical language modeling, and machine learning.
As a senior principal data scientist for Honeywell, Tonya developed machine-learned predictive models to improve voice-controlled and other Internet of Things and aerospace products. Her earlier efforts as a senior scientist for eBay focused on search results ranking, query parsing, inventory prediction, catalog attribution and product classification. As a member of Thomson Reuters Research & Development from 2004 to 2012, Tonya worked on numerous projects, including Concept Search, non-English stemming and decompounding, document recommendation systems, question answering, and named entity tagging.