NASSLLI 2012 June 18 - 22

Sign up here for our Mailing List:

Questions? Email us:

Search the site:

  Powered by Google

Extracting Social Meaning and Sentiment

Room: UTC 4.104


The course will review central results from linguistics and cognitive psychology concerning the dimensions of affectivity and the ways in which attitudinal and emotional information is expressed in language. It will then focus on the computational extraction of such information from textual data using supervised and unsupervised methods. Special emphasis will be placed on bringing social and contextual information into the models and designing features that capture non-literal uses of language.


None, though students with programming skills will be happier and more productive in the class. Students can write to me with tips about how to prepare effectively.


Students could study these materials and play around with the demos:


  1. Background and goals: the linguistic and computational challenges of sentiment and social meaning; applications of sentiment analysis in academia and industry
  2. Introduction to course data sets; text preparation for sentiment analysis (tokenizing, parsing, contextual features)
  3. Supervised methods for sentiment analysis; the importance of modeling the blended, continuous nature of sentiment
  4. Characterizing and modeling non-literal language
  5. Unsupervised methods for sentiment; wrap-up


Christopher Potts



Christopher Potts is Associate Professor of Linguistics at Stanford University. In his research, he uses computational methods to explore how emotion is expressed in language and how linguistic production and interpretation are influenced by the context of utterance. He is the author of the 2005 book The Logic of Conventional Implicatures, as well as numerous scholarly papers in computational and theoretical linguistics. In his free time, he enjoys running, cycling, and skateboarding.