- Mathematics and Computer Science
Amber Stubbs graduated from Simmons University in 2005 with an B.S. in Computer Science and English, then attended Brandeis University where she earned her Ph.D. in Computer Science, specifically in the field of natural language processing. While at Brandeis she co-authored a book, Natural Language Annotation for Machine Learning (O’Reilly, 2012) with James Pustejovksy.
Her doctoral dissertation, "A Methodology for Using Professional Knowledge in Corpus Annotation," involved creating an annotation methodology to extract high-level information — such a hospital patient's medical diagnosis — from narrative texts. As part of that research, she also developed the Multi-Purpose Annotation Environment (MAE) and Multi-document Adjudication Interface (MAI) software, which is used at institutions around the world for natural language processing research.
After completing her Ph.D., Amber worked as a Postdoctoral Associate under Ozlem Uzuner at the State University of New York at Albany. During that time, she worked on the 2014 i2b2 Natural Language Processing Shared Task, which focused on recognizing risk factors for heart disease in medical records, as well as the identification and removal of personal information about patients from their records.
Amber became an Assistant Professor at Simmons University in 2014, where she teaches courses in both the Computer Science and LIS programs. She is delighted to be back at Simmons, and enjoys helping students understand technology and how to make it work for them.
What I Teach
- CS232: Data Structures and Algorithms
- CS330: Structure and Organization of Programming Languages
- LIS 488: Technology for Information Professionals
Stubbs' research is primarily in natural language processing, particularly of medical records. She experiments with ways that the information which medical professionals posses can be collected and formatted so that computers can understand and use that information to augment current health practices. She is especially interested in best practices for the annotation all types of high-level information, not limited to medical records.