Abstract
We explore the commonly stated assumption that the standard POMDP formalism for belief updates cannot be directly applied to Dialogue Management for Spoken Dialogue Systems (SDSs) due to the computational intractability of maintaining a large belief state space. Focusing on SDSs, as this application has particular bounds in terms of “real-time” belief updates and potentially massive numbers of observations, we quantify computational constraints both in terms of compute time and memory. We establish a level of complexity of SDS task below which a direct implementation of the standard POMDP formalism is possible and beyond which some form of compressed representation is required. We find that computation time of POMDP belief updates is rarely an issue. Memory size and latency tend to be the dominant constraints. Low-latency, shared-memory architectures are more suitable than General Purpose Graphics Processing Units (GPGPUs) or largescale cluster/cloud infrastructure. One assumption, that users do not change their goal during a dialogue, has significant beneficial impacts on memory requirements allowing for practical POMDP SDSs which have millions of states.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Black, A.W., Burger, S., Conkie, A., Hastie, H., Keizer, S., Lemon, O., Merigaud, N., Parent, G., Schubiner, G., Thomson, B., Williams, J.D., Yu, K., Young, S., Eskenazi, M.: Spoken Dialog Challenge 2010: Comparison of Live and Control Test Results. In: Prc. SIGdial (2011)
Bull, M., Aylett, M.: An analysis of the timing of turn-taking in a corpus of goal-oriented dialogue. In: Proceeding of ISCLP (1998)
Chandramohan, S., Geist, M., Pietquin, O.: Sparse approximate dynamic programming for dialog management. In: Proceedings of SIGdial (2010)
Crook, P.A., Lemon, O.: Representing uncertainty about complex user goals in statistical dialogue systems. In: Proceedings of SIGdial (2010)
Crook, P.A., Lemon, O.: Lossless Value Directed Compression of Complex User Goal States for Statistical Spoken Dialogue Systems. In: Proceedings of Interspeech (2011)
Gašić, M., Young, S.: Effective Handling of Dialogue State in the Hidden Information State POMDP-based Dialogue Manager. ACM Transactions in Speech and Language Processing 7(3) (2011)
Henderson, J., Lemon, O.: Mixture Model POMDPs for Efficient Handling of Uncertainty in Dialogue Management. In: Proceedings of ACL (2008)
Lemon, O., Georgila, K., Henderson, J.: Evaluating Effectiveness and Portability of Reinforcement Learned Dialogue Strategies with real users: the TALK TownInfo Evaluation. In: IEEE/ACL Spoken Language Technology, pp. 178– 181 (2006)
Möller, S.: Assessment and Prediction of Speech Quality in Telecommunications. Kluwer (2000)
Möller, S., Engelbrecht, K.P., Schleicher, R.: Predicting the quality and usability of spoken dialogue services. Speech Commincation 50, 730–744 (2008)
Raux, A.: Flexible turn-taking for spoken dialog systems. Ph.D. thesis, CMU (2008)
Spaan, M., Vlassis, N.: Perseus: randomized point-based value iteration for POMDPs. Tech. Rep. IAS-UVA-04-02, Universiteit van Amsterdam (2004)
Thomson, B., Young, S.: Bayesian update of dialogue state: a POMDP framework for spoken dialogue systems. Computer Speech and Language 24(4), 562–588 (2010)
Williams, J., Poupart, P., Young, S.: Factored partially observable markov decision processes for dialogue management. In: Workshop on Knowledge and Reasoning in Practical Dialog Systems (IJCAI) (2005)
Williams, J., Young, S.: Partially Observable Markov Decision Processes for Spoken Dialog Systems. Computer Speech and Language 21(2), 231–422 (2007)
Williams, J.D.: Incremental partition recombination for efficient tracking of multiple dialog states. In: Proceeding of ICASSP, pp. 5382 – 5385 (2010)
Williams, J.D., Young, S.: Scaling Up POMDPs for Dialog Management: The ”Summary POMDP” Method. In: Proceedings of ASRU (2005)
Young, S., Gašić, M., Keizer, S., Mairesse, F., Thomson, B., Yu, K.: The Hidden Information State model: a practical framework for POMDP based spoken dialogue management. Computer Speech and Language 24(2), 150–174 (2010)
Acknowledgements
The authors would like to thank the Engineering and Physical Sciences Research Council, UK (EPSRC) grant number EP/G069840/1, and partial funding from the European Community’s Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 270019 (SPACEBOOK project www.spacebook-project.eu).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2011 Springer Science+Business Media, LLC
About this paper
Cite this paper
Crook, P.A., Roblin, B., Loidl, HW., Lemon, O. (2011). Parallel Computing and Practical Constraints when applying the Standard POMDP Belief Update Formalism to Spoken Dialogue Management. In: Delgado, RC., Kobayashi, T. (eds) Proceedings of the Paralinguistic Information and its Integration in Spoken Dialogue Systems Workshop. Springer, New York, NY. https://doi.org/10.1007/978-1-4614-1335-6_20
Download citation
DOI: https://doi.org/10.1007/978-1-4614-1335-6_20
Published:
Publisher Name: Springer, New York, NY
Print ISBN: 978-1-4614-1334-9
Online ISBN: 978-1-4614-1335-6
eBook Packages: EngineeringEngineering (R0)