Brendan O'Connor

Email: brenocon@cs.umass.edu
Twitter: @brendan642

Assistant Professor, School of Computer Science, University of Massachusetts Amherst
Core Faculty, Computational Social Science Initiative
Office: Room 222, Computer Science Building, 140 Governors Drive (Amherst, MA 01003-9264)
Map with directions (10 MB pdf), google maps, campus map


I recently joined UMass Amherst's School of Computer Science as an assistant professor, where I am also a Core Faculty affiliate of the Computational Social Science Initiative. Within CS, I am also affiliated with CIIR and MLDS.

Teaching:
This fall at UMass, I will be teaching CMPSCI 585, Introduction to Natural Language Processing.

Research:
What can statistical text analysis tell us about society? I develop text mining methods that can help answer social science questions. I'm interested in statistical machine learning and natural language processing, especially when informed by or applied to areas like political science or sociolinguistics. My work often uses text data from news and social media. Some areas of interest include:

  • Interactive text data visualization (prev work)
  • Regional dialects and language change in social media (prev work (1) (2) (3))
  • Social media sentiment compared to polls (prev work)
  • NLP for informal (sometimes called "noisy") language (prev work)
  • Syntactic analysis, e.g. part of speech tagging, parsing
  • Semantic/discourse analysis, e.g. coreference, events, entity modeling (e.g. movie character archetypes)
  • Political event extraction from news (prev work)
  • Joint statistical models of social factors and language
  • Measuring sentiment, opinions, ideologies, worldviews from text
  • Human-in-the-loop learning, e.g. simple syntax annotations, or Mechanical Turk crowdsourcing
  • Probabilistic graphical models, latent variable models, Bayesian inference, sampling and optimization methods

See also my (oldish) research statement or publications below. If you are interested in getting involved in research, shoot me an email. Keep in mind there is a rich set of faculty at UMass interested in similar areas—from computational social science to natural language processing—including but not limited to James Kitts (Soc), Bruce Desmarais (Polisci), Krista Gile (Stats), David Jensen (CS), Hanna Wallach (CS), Bruce Croft (CS), James Allan (CS), Andrew McCallum (CS), Brian Dillon (Ling), Kristine Yu (Ling), Rajesh Bhatt (Ling), and many more! See also the CSSI website.

Background:
My PhD was in the Machine Learning Department in Carnegie Mellon University's School of Computer Science, where I was a member of the Noah's ARK research group. I have also been a Visiting Fellow at Harvard IQSS, and an intern on the Facebook Data Science team. Before grad school, I worked on crowdsourced annotations at CrowdFlower / Dolores Labs, as well as "semantic" search at Powerset. I was an undergrad and masters student in the Stanford Symbolic Systems Program (cognitive science, more or less).

Recent stuff

Videos from past presentations

Selected recent publications

(See also Google Scholar.)

Other publications

Other papers on my CV or Google Scholar.

Software

  • MiTextExplorer: interactive exploration of text data and document covariates.
  • TweetNLP: tokenization and part-of-speech tagging for Twitter.
  • ARKref, a coreference resolution system.
  • ParseViz - quick and dirty parse tree/dependency visualization via graphviz.
  • tsvutils for tab-separated data processing
  • Other misc utilities (commandline, R, Python...)

Recent and not-so-recent news

Demos etc.

  • Inactive demos

    Elsewhere on the Internet

  • Twitter
  • Facebook
  • LinkedIn
  • Github
  • Gists
  • Writings:

    My PGP key

    Other O'Connors

    There are many Brendan O'Connors in the world. If this is the wrong webpage, you may be interested in another Brendan O'Connor; for example, My awesome sister, Maureen O'Connor, is a writer in New York.