Article Structure
Abstract
This work presents an agenda-based approach to improve the robustness of the dialog manager by using dialog examples and n-best recognition hypotheses.
Introduction
Development of spoken dialog systems involves human language technologies which must cooperate to answer user queries.
Related Work
In many spoken dialog systems that have been developed recently, various knowledge sources are used.
Example-based Dialog Modeling
Our approach is implemented based on Example-Based Dialog Modeling (EBDM) which is one of generic dialog modelings.
Agenda Graph
In this paper, agenda graph G is simply a way of encoding the domain-specific dialog control to complete the task.
Greedy Selection with n-best Hypotheses
Many speech recognizers can generate a list of plausible hypotheses (n-best list) but output only the most probable one.
Error Recovery Strategy
As noted in Section 4.2, the discourse interpretation sometimes fails to generate candidate nodes.
Experiment & Result
First we developed the spoken dialog system for PH OPE in which an intelligent robot can provide information about buildings (i.e., room number, room location, room name, room type) and people (i.e., name, phone number, email address, cellular phone number).
Conclusion & Discussion
This paper has proposed a new agenda-based approach with n-best recognition hypotheses to improve the robustness of the Example-based Dialog Modeling (EBDM) framework.
Topics
score functions
- Given the agenda graph and n-best hypotheses, the system can predict the next system actions to maximize multilevel score functions .Page 1, “Abstract”
- the score function based on current input and discourse structure given the focus stack.Page 5, “Agenda Graph”
- Therefore, we need to select the hypothesis that maximizes the scoring function among a set of n-best hypotheses of each utterance.Page 5, “Greedy Selection with n-best Hypotheses”
- Secondly, the multilevel score functions are computed for each candidate node Ci given a hypothesis hi.Page 5, “Greedy Selection with n-best Hypotheses”
- Otherwise, the best node which would be pushed onto the focus stack must be selected using multilevel score functions .Page 6, “Greedy Selection with n-best Hypotheses”
- The node selection is determined by calculating some score functions .Page 6, “Greedy Selection with n-best Hypotheses”
- We defined multilevel score functions that combine the scores of ASR, SLU, and DM modules, which range from 0.00 to 1.00.Page 6, “Greedy Selection with n-best Hypotheses”
- For the node selection, we divided the score function into two functions 8H0”), hypothesis score, and SD (ci|S), discourse score, where Ci is the focus node to be generated by single hypothesis hi.Page 6, “Greedy Selection with n-best Hypotheses”
See all papers in Proc. ACL 2008 that mention score functions.
See all papers in Proc. ACL that mention score functions.
Back to top.
discourse structure
- The focus stack takes into account the discourse structure by keeping track of discourse states.Page 4, “Agenda Graph”
- the score function based on current input and discourse structure given the focus stack.Page 5, “Agenda Graph”
- In addition to the hypothesis score, we defined the discourse score SD at the discourse level to consider the discourse structure between the previous node and current node given the focus stack 8.Page 6, “Greedy Selection with n-best Hypotheses”
See all papers in Proc. ACL 2008 that mention discourse structure.
See all papers in Proc. ACL that mention discourse structure.
Back to top.
natural language
- Since the performance in human language technologies such as Automatic Speech Recognition (ASR) and Natural Language Understanding (NLU)1 have been improved, this advance has made it possible to develop spoken dialog systems for many different application domains.Page 1, “Introduction”
- 1Through this paper, we will use the term natural language to include both spoken language and written languagePage 1, “Introduction”
- The EBDM framework is a simple and powerful approach to rapidly develop natural language interfaces for multi-domain dialog processing (Lee et al., 2006b).Page 3, “Example-based Dialog Modeling”
See all papers in Proc. ACL 2008 that mention natural language.
See all papers in Proc. ACL that mention natural language.
Back to top.
