DMCA
Partially observable markov decision processes with continuous observations for dialogue management (2005)
Cached
Download Links
| Venue: | Computer Speech and Language |
| Citations: | 217 - 52 self |
BibTeX
@INPROCEEDINGS{Williams05partiallyobservable,
author = {Jason D. Williams},
title = {Partially observable markov decision processes with continuous observations for dialogue management},
booktitle = {Computer Speech and Language},
year = {2005},
pages = {393--422}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
This work shows how a dialogue model can be represented as a Partially Observable Markov Decision Process (POMDP) with observations composed of a discrete and continuous component. The continuous component enables the model to directly incorporate a confidence score for automated planning. Using a testbed simulated dialogue management problem, we show how recent optimization techniques are able to find a policy for this continuous POMDP which outperforms a traditional MDP approach. Further, we present a method for automatically improving handcrafted dialogue managers by incorporating POMDP belief state monitoring, including confidence score information. Experiments on the testbed system show significant improvements for several example handcrafted dialogue managers across a range of operating conditions. 1
Keyphrases
dialogue management observable markov decision continuous observation continuous component dialogue manager dialogue model automated planning partially observable markov decision process confidence score information dialogue management problem recent optimization technique pomdp belief state monitoring traditional mdp approach several example operating condition testbed system show significant improvement confidence score continuous pomdp
