Presentation 2007/12/13
Optimization of Question and Answer Database for Speech-oriented Guidance System : Response Generation Based on Question and Answer Database Using Automatic Speech Recognition Results and Its Optimization
Shota TAKEUCHI, Tobias CINCAREK, Hiromichi KAWANAMI, Hiroshi SARUWATARI, Kiyohiro SHIKANO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) A speech-oriented dialog system employed in real-environment requires dialog control techniques which enable flexible processing of various utterance expressions. As a robust response generation method for various utterance expressions, an example-based response generation method has been studied. This method employs a question and answer database (QADB) to generate the most appropriate response by selecting an example question which is nearest to an input. Conventionally, a QADB is constructed from transcriptions of utterance data. In this paper, we propose response generation based on a question and answer database using automated speech recognition results (ASR-QADB),which enables to perform robust selection of examples against recognition errors. Evaluating the performance of system responses by response accuracy, the same response accuracy as with the conventional method is achieved by constructing and optimizing ASR-QADB using 6-best recognition results.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) speech-oriented dialog system / example-based response generation / QADB optimization / leave-one-out cross-validation / QADB using automated speech recognition results
Paper # NLC2007-83,SP2007-146
Date of Issue

Conference Information
Committee NLC
Conference Date 2007/12/13(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Natural Language Understanding and Models of Communication (NLC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Optimization of Question and Answer Database for Speech-oriented Guidance System : Response Generation Based on Question and Answer Database Using Automatic Speech Recognition Results and Its Optimization
Sub Title (in English)
Keyword(1) speech-oriented dialog system
Keyword(2) example-based response generation
Keyword(3) QADB optimization
Keyword(4) leave-one-out cross-validation
Keyword(5) QADB using automated speech recognition results
1st Author's Name Shota TAKEUCHI
1st Author's Affiliation Graduate School of Information Science, Nara Institute of Science and Technology()
2nd Author's Name Tobias CINCAREK
2nd Author's Affiliation Graduate School of Information Science, Nara Institute of Science and Technology
3rd Author's Name Hiromichi KAWANAMI
3rd Author's Affiliation Graduate School of Information Science, Nara Institute of Science and Technology
4th Author's Name Hiroshi SARUWATARI
4th Author's Affiliation Graduate School of Information Science, Nara Institute of Science and Technology
5th Author's Name Kiyohiro SHIKANO
5th Author's Affiliation Graduate School of Information Science, Nara Institute of Science and Technology
Date 2007/12/13
Paper # NLC2007-83,SP2007-146
Volume (vol) vol.107
Number (no) 405
Page pp.pp.-
#Pages 6
Date of Issue