Brendan O'Connor

Brendan's Photo
Assistant Professor
348 CS Building
(413) 577-2503


Natural language processing, computational social science, statistical machine learning, social media.


What can statistical text analysis tell us about society?  Prof. O'Connor develops natural language processing and machine learning methods to help scientific investigation about political and social trends.  For example, he and his colleagues have analyzed Twitter to understand how new slang spreads between cities, how textual sentiment corresponds to public opinion polls, censorship in Chinese microblogs, and how to extract events in international relations from the news. Statistical language patterns can give insight into the underlying social variables (text as measurement); or, they can reveal the socially embedded process of language generation.  He is affiliated with research groups for data science, information retrieval, cognitive science, and computational social science at UMass.


Ph.D., Machine Learning, Carnegie Mellon University (2014); B.S. and M.S., Symbolic Systems, Stanford University (2006). From 2006 to 2009, Professor O'Connor worked at the technology startups Powerset and Crowdflower / Dolores Labs. During his PhD studies, he also interned on the Facebook Data Science team, and visited Harvard's Institute for Quantitative Social Science.