Abstract
Using the techniques of Chap. 3 one can build a working spoken dialogue system for some limited real-world situations. There are still, however, significant computational difficulties when dealing with large state spaces. In particular, the Loopy Belief Propagation algorithm does not scale well when individual nodes have very large numbers of values. This chapter will discuss methods for improving the tractability of the standard algorithm.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
Note that the HIS model does allow for some extra features, such as prior information when partitions are split.
- 2.
All experiments were performed on an Intel Core 2 Quad CPU, 2.4 GHz processor with 4 GB of RAM.
References
Lasserre J, Kannan A, Winn J (2007) Hybrid learning of large jigsaws. In: Proceedings of CVPR, pp 1–8
Minka T (2001) Expectation propagation for approximate Bayesian inference. In: Proceedings of uncertainty in AI, pp 362–369
Siddiqi SM, Moore AW (2005) Fast inference and learning in large-state-space HMMs. In: Proceedings of ICML, 2005
Young S, Gasic M, Keizer S, Mairesse F, Schatzmann J, Thomson B, Yu K (2009) The hidden information state model: a practical framework for POMDP-based spoken dialogue management. Comput Speech Lang (2009). ISSN 08852308
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2013 Springer-Verlag London
About this chapter
Cite this chapter
Thomson, B. (2013). Maintaining State: Optimisations. In: Statistical Methods for Spoken Dialogue Management. Springer Theses. Springer, London. https://doi.org/10.1007/978-1-4471-4923-1_4
Download citation
DOI: https://doi.org/10.1007/978-1-4471-4923-1_4
Published:
Publisher Name: Springer, London
Print ISBN: 978-1-4471-4922-4
Online ISBN: 978-1-4471-4923-1
eBook Packages: EngineeringEngineering (R0)