Evaluation of a Large Language Model to Identify Confidential Content in Adolescent Encounter Notes
Naveed Rabbani, Conner Brown, Michael Bedgood, Rachel Goldstein, Jennifer L. Carlson, Natalie M. Pageler, Keith Morse
Abstract
This cross-sectional study assesses the ability of a language learning model to classify whether a progress note contains confidential information and to identify the specific confidential content in the note.
Topics & Concepts
MedicineConfidentialityContent (measure theory)Natural language processingComputer securityMathematical analysisMathematicsComputer scienceTopic ModelingMisinformation and Its ImpactsSpeech Recognition and Synthesis