We describe research on endowing spoken language systems with the ability to consider the cost of misrecognition, and using that knowledge to guide clarification dialog about a users intentions. Our approach relies on coupling utility-directed policies for dialog with the ongoing Bayesian fusion of evidence obtained from multiple utterances recognized during an interaction. After describing the methodology, we review the operation of a prototype system called DeepListener. DeepListener considers evidence gathered about utterances over time to make decisions about the optimal dialog strategy or realworld action to take given uncertainties about a users intentions and the costs and benefits of different outcomes.
Cite as: Horvitz, E., Paek, T. (2000) Deeplistener: harnessing expected utility to guide clarification dialog in spoken language systems. Proc. 6th International Conference on Spoken Language Processing (ICSLP 2000), vol. 1, 226-229, doi: 10.21437/ICSLP.2000-56
@inproceedings{horvitz00_icslp,
author={Eric Horvitz and Tim Paek},
title={{Deeplistener: harnessing expected utility to guide clarification dialog in spoken language systems}},
year=2000,
booktitle={Proc. 6th International Conference on Spoken Language Processing (ICSLP 2000)},
pages={vol. 1, 226-229},
doi={10.21437/ICSLP.2000-56}
}