Can I finish? Learning when to respond to incremental interpretation results in interactive dialogue

David DeVault, Kenji Sagae and David Traum

SIGDIAL Workshop on Discourse and Dialogue (SIGDIAL 2009)
Queen Mary University of London, September 11-12, 2009

Summary

We investigate novel approaches to responsive overlap behaviors in dialogue systems, opening possibilities for virtual humans to interrupt, acknowledge or complete a user's utterance while it is in progress. Our specific contributions are a method for determining whether the system's understanding of an utterance has reached a point suitable for completions or interruptions, and a prototype implementation of agent completions of user utterances within a mature dialogue system. This framework facilitates the implementation of a wide range of behaviors that are common in human dialogue, but have been largely absent in dialogue systems.