Abstract | ||
---|---|---|
Recent research has shown that effective dialogue management can be achieved through the Partially Observable Markov Decision Process (POMDP) framework. However past research on POMDP-based dialogue systems usually assumed the parameters of the decision process were known a priori. The main contribution of this paper is to present a Bayesian reinforcement learning framework for learning the POMDP parameters online from data, in a decision-theoretic manner. We discuss various approximations and assumptions which can be leveraged to ensure computational tractability, and apply these techniques to learning observation models for several simulated spoken dialogue domains. |
Year | DOI | Venue |
---|---|---|
2012 | 10.1109/JSTSP.2012.2229962 | J. Sel. Topics Signal Processing |
Keywords | Field | DocType |
partially observable markov decision process,speech recognition,bayesian inference,reinforcement learning,adaptive dialogue system,bayesian reinforcement learning,bayes methods,learning (artificial intelligence),dialogue management,spoken dialogue domain,computational tractability,pomdp,partially observable markov decision process (pomdp),markov decision process (mdp),decision theory,interactive systems,markov processes,learning artificial intelligence | Markov process,Markov model,Partially observable Markov decision process,Computer science,A priori and a posteriori,Markov decision process,Artificial intelligence,Decision theory,Decision process,Machine learning,Bayes' theorem | Journal |
Volume | Issue | ISSN |
6 | 8 | 1932-4553 |
Citations | PageRank | References |
5 | 0.43 | 31 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Shao Wei Png | 1 | 117 | 5.88 |
Joelle Pineau | 2 | 2857 | 184.18 |
Chaib-draa, Brahim | 3 | 1190 | 113.23 |