Days and Times: Tu 9:00am-11:30am
Room: 4430 Posvar Hall
Instructor: Michael Colaresi
This is a pre-approved elective for Spring Term 2019.

The amount of social science relevant information stored in text presses against the boundaries of human comprehension. From corpora of international and national speeches and documents, through repositories of human rights reports, news archives and blog posts, to streaming social media feeds, a font of knowledge awaits those that have the creativity and ability to model and learn from text as data. This class is a lower case, but more than cursory, introduction to the use of natural language processing, machine learning and Bayesian inference frameworks and tools to accelerate innovations in key social science research. The workload is significant but should not give anyone a stroke. We will cover lexical methods such as dictionary-based sentiment analysis, topic modeling and more general supervised learning techniques, as well as research designs that leverage syntactic information. Theoretical lectures will be punctuated with hands-on coding assignments. The point of the class is to increase student's confidence in utilizing text in their applied work.