Loading...

Using Partially-Observable Markov Decision Process for Dialogue Management in Spoken Dialogue Systems

Rahbar Noudehi, Siavash | 2010

535 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 40288 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Sameti, Hossein
  7. Abstract:
  8. The use of Spoken Dialogue Systems is growing everyday and these systems will substitute current Iterative Voice Response systems in near future. A Spoken Dialogue System consists of Speech Recognition, Language Understanding, Dialogue Management, Speech Generation and Text to Speech Modules. Among these modules the only one that is specific part of Dialogue Systems is Dialogue Management. The responsibility of this part is to determine system behavior to maximize specific variables such as user goal finding accuracy and speed of finding the goal. There were different approaches to dialogue management in recent years the use of Partially-Observable Markov Decision Processes was very popular among them. In this thesis use of POMDP for a Bank Information Application and reducing a Menu Selection Problem to a POMDP problem has been discussed. The main problem in using POMDPs is the computational complexity of solutions to it. To solve this problem, a new method has been provided based on dividing the POMDP problem to smaller problems and solving them, then merging acquired policies to create a policy for the main problem. Desired properties for the optimal policy have been declared that makes this method to be able to return it. Then we discussed why these properties are present in a menu selection problem. We also implemented other modules of the spoken dialogue system to make the system usable in practice
  9. Keywords:
  10. Spoken Dialoge System ; Partially Observable Markov Decision Process ; Dialogue Management

 Digital Object List

  • محتواي پايان نامه
  •   view

 Bookmark

No TOC