Faculty Expert Profile

Mark Dredze, a computer science professor, is internationally recognized for mining social media data to pioneer new applications in public health informatics. One of the founders of this emerging field, Dredze develops machine learning and natural language processing tools to gain insights from the web's endless data to better understand human behavior and to inform public health policy and interventions. For instance, he used Twitter data to track the spread of the flu with accuracy comparable to the CDC.

His research is providing unprecedented insights on suicide prevention, vaccine refusal, HIV, tobacco, gun violence, and other public health issues. In 2017, he and Michael Paul at the University of Colorado co-authored "Social Monitoring for Public Health"—the first book surveying this nascent field.

Dredze develops new methods based on machine learning for a wide range of natural language processing tasks, including information extraction, semantics, speech processing, and topic modeling. He is one of the leaders at Johns Hopkins University and Hospital in mining the free text found in clinical patient records to support patient care and discover population trends. He collaborates broadly with clinicians in the Johns Hopkins School of Medicine and the Bloomberg School of Public Health.