Reinforcement Learning for Spoken Dialogue Systems

Abstract

Recently, a number of authors have proposed treating dialogue systems as Markov decision processes (MDPs). However, the practical application ofMDP algorithms to dialogue systems faces a number of severe technical challenges. We have built a general software tool (RLDS, for Reinforcement Learning for Dialogue Systems) based on the MDP framework, and have applied it to dialogue corpora gathered from two dialogue systems built at AT&T Labs. Our experiments demonstrate that RLDS holds promise as a tool for "browsing" and understanding correlations in complex, temporally dependent dialogue corpora.

Cite

Text

Singh et al. "Reinforcement Learning for Spoken Dialogue Systems." Neural Information Processing Systems, 1999.

Markdown

[Singh et al. "Reinforcement Learning for Spoken Dialogue Systems." Neural Information Processing Systems, 1999.](https://mlanthology.org/neurips/1999/singh1999neurips-reinforcement/)

BibTeX

@inproceedings{singh1999neurips-reinforcement,
  title     = {{Reinforcement Learning for Spoken Dialogue Systems}},
  author    = {Singh, Satinder P. and Kearns, Michael J. and Litman, Diane J. and Walker, Marilyn A.},
  booktitle = {Neural Information Processing Systems},
  year      = {1999},
  pages     = {956-962},
  url       = {https://mlanthology.org/neurips/1999/singh1999neurips-reinforcement/}
}