INFO 4350 - Conversations and Information

Fall. 3 credits. Letter Grades Only.

Prerequisites: INFO 4300 or CS 4300, MATH 2940 or equivalent, CS 2800 or equivalent and CS 2110 or equivalent. Permission of instructor is required.

C. Danescu-Niculescu-Mizil.

As conversations become central to the way in which we acquire and consume information and as the societal implications became evident (e.g., potential for misinformation, hate speech, ethical questions surrounding the use of chatbots), there is an increasing need for automated ways to analyze their quality. This course will cover computational approaches to conversational analysis: starting from topics that were traditionally addressed within the Sociology subfield of Conversational Analysis, continuing with reviewing existing computational methods that address those topics, and further guiding the development of new methods and their application to real-life conversational data. In addition to lecturing in-class activities will include close interaction with conversational data, such as collaborative annotation of conversational phenomen, in-depth discussion of hand-picked examples, as well as joint and iterative development of annotation guidelines that lend themselves for computational modeling.

Outcome 1: The students will be able to identify and characterize conversational phenomena

Outcome 2: The students will be able to analyze and annotate conversational data

Outcome 3: The students will be able to analyze conversations using computational techniques