Paper: Using Emotion to Gain Rapport in a Spoken Dialog System

ACL ID N09-3009
Title Using Emotion to Gain Rapport in a Spoken Dialog System
Venue HLT-NAACL Companion Volume: Student Research Workshop and Doctoral Consortium
Session
Year 2009
Authors

This paper describes research on automati- cally building rapport. This is done by adapt- ing responses in a spoken dialog system to users’ emotions as inferred from nonverbal voice properties. Emotions and their acous- tic correlates will be extracted from a persua- sive dialog corpus and will be used to imple- ment an emotionally intelligent dialog system; one that can recognize emotion, choose an op- timal strategy for gaining rapport, and render a response that contains appropriate emotion, both lexically and auditory. In order to deter- mine the value of emotion modeling for gain- ing rapport in a spoken dialog system, the final implementation will be evaluated using differ- ent configurations through a user study.