Title: Detecting Public Health Information in General Text

Abstract:
Privacy is a growing concern of the general public, and there is concern about unauthorized disclosure and use of an individual's personal health information (PHI).  In order to secure this data (by encryption, anonymization, or some other means), the relevant data must first be identified.  The presentation will focus on how PHI can be detected in general texts using natural language processing methods and a number of specialized datasets.  The results from an initial experiment will also be presented.