Improving identification accuracy by extending acceptable utterances in spoken dialogue system using barge-in timing

Kyoko Matsuyama*, Kazunori Komatani, Toru Takahashi, Tetsuya Ogata, Hiroshi G. Okuno

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We describe a novel dialogue strategy enabling robust interaction under noisy environments where automatic speech recognition (ASR) results are not necessarily reliable. We have developed a method that exploits utterance timing together with ASR results to interpret user intention, that is, to identify one item that a user wants to indicate from system enumeration. The timing of utterances containing referential expressions is approximated by Gamma distribution, which is integrated with ASR results by expressing both of them as probabilities. In this paper, we improve the identification accuracy by extending the method. First, we enable interpretation of utterances including ordinal numbers, which appear several times in our data collected from users. Then we use proper acoustic models and parameters, improving the identification accuracy by 4.0% in total. We also show that Latent Semantic Mapping (LSM) enables more expressions to be handled in our framework.

Original languageEnglish
Title of host publicationTrends in Applied Intelligent Systems - 23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2010, Proceedings
Pages585-594
Number of pages10
EditionPART 2
DOIs
Publication statusPublished - 2010
Externally publishedYes
Event23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligence Systems, IEA/AIE 2010 - Cordoba, Spain
Duration: 2010 Jun 12010 Jun 4

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 2
Volume6097 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligence Systems, IEA/AIE 2010
Country/TerritorySpain
CityCordoba
Period10/6/110/6/4

Keywords

  • barge-in
  • conversational interaction
  • spoken dialogue systems
  • utterance timing

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'Improving identification accuracy by extending acceptable utterances in spoken dialogue system using barge-in timing'. Together they form a unique fingerprint.

Cite this