Language Science Lunch Talk - Modeling Vocal Entrainment: Automated and Experimental Approaches
Date:
Invited talk at the Language Science Center, University of Maryland
Abstract
When people engage in conversations, cooperative tasks, or information-exchange, they tend to unconsciously align their linguistic features. This entrainment has been observed across the vocal (or acoustic-prosodic), lexical, semantic, and syntactic levels, and is indicative of the quality of interaction and success in cooperative tasks.
This talk will focus on modeling vocal entrainment with large dyadic conversational data, and will share findings on predicting cohesion in two-party conversations. It will also shed light on the impact of group size and complex turn-taking on vocal entrainment, and discuss the inherent challenges of modeling entrainment in multi-party spoken conversations.
Permalink:https://languagescience.umd.edu/events/language-science-lunch-talk-megh-krishnaswamy