A graph-based spoken dialog strategy utilizing multiple understanding hypotheses

Norihide Kitaoka, Yuji Kinoshita, Sunao Hara, Chiyomi Miyajima, Kazuya Takeda

Research output: Contribution to journalArticle

Abstract

We regarded a dialog strategy for information retrieval as a graph search problem and proposed several novel dialog strategies that can recover from misrecognition through a spoken dialog that traverses the graph. To recover from misrecognition without seeking confirmation, our system kept multiple understanding hypotheses at each turn and searched for a globally optimal hypothesis in the graph whose nodes express understanding states across user utterances in a whole dialog. In the search, we used a new criterion based on efficiency in information retrieval and consistency with understanding hypotheses, which is also used to select an appropriate system response. We showed that our system can make more efficient and natural dialogs than previous ones.

Original languageEnglish
Pages (from-to)1-10
Number of pages10
JournalTransactions of the Japanese Society for Artificial Intelligence
Volume29
Issue number1
DOIs
Publication statusPublished - 2014

Fingerprint

Information retrieval

Keywords

  • Consistency measure
  • Efficiency measure
  • Misrecognition
  • Response generation
  • Spoken dialog strategy

ASJC Scopus subject areas

  • Artificial Intelligence
  • Software

Cite this

A graph-based spoken dialog strategy utilizing multiple understanding hypotheses. / Kitaoka, Norihide; Kinoshita, Yuji; Hara, Sunao; Miyajima, Chiyomi; Takeda, Kazuya.

In: Transactions of the Japanese Society for Artificial Intelligence, Vol. 29, No. 1, 2014, p. 1-10.

Research output: Contribution to journalArticle

Kitaoka, Norihide ; Kinoshita, Yuji ; Hara, Sunao ; Miyajima, Chiyomi ; Takeda, Kazuya. / A graph-based spoken dialog strategy utilizing multiple understanding hypotheses. In: Transactions of the Japanese Society for Artificial Intelligence. 2014 ; Vol. 29, No. 1. pp. 1-10.
@article{b1106d50320f461d8b994bf28bbc7929,
title = "A graph-based spoken dialog strategy utilizing multiple understanding hypotheses",
abstract = "We regarded a dialog strategy for information retrieval as a graph search problem and proposed several novel dialog strategies that can recover from misrecognition through a spoken dialog that traverses the graph. To recover from misrecognition without seeking confirmation, our system kept multiple understanding hypotheses at each turn and searched for a globally optimal hypothesis in the graph whose nodes express understanding states across user utterances in a whole dialog. In the search, we used a new criterion based on efficiency in information retrieval and consistency with understanding hypotheses, which is also used to select an appropriate system response. We showed that our system can make more efficient and natural dialogs than previous ones.",
keywords = "Consistency measure, Efficiency measure, Misrecognition, Response generation, Spoken dialog strategy",
author = "Norihide Kitaoka and Yuji Kinoshita and Sunao Hara and Chiyomi Miyajima and Kazuya Takeda",
year = "2014",
doi = "10.1527/tjsai.29.1",
language = "English",
volume = "29",
pages = "1--10",
journal = "Transactions of the Japanese Society for Artificial Intelligence",
issn = "1346-0714",
publisher = "Japanese Society for Artificial Intelligence",
number = "1",

}

TY - JOUR

T1 - A graph-based spoken dialog strategy utilizing multiple understanding hypotheses

AU - Kitaoka, Norihide

AU - Kinoshita, Yuji

AU - Hara, Sunao

AU - Miyajima, Chiyomi

AU - Takeda, Kazuya

PY - 2014

Y1 - 2014

N2 - We regarded a dialog strategy for information retrieval as a graph search problem and proposed several novel dialog strategies that can recover from misrecognition through a spoken dialog that traverses the graph. To recover from misrecognition without seeking confirmation, our system kept multiple understanding hypotheses at each turn and searched for a globally optimal hypothesis in the graph whose nodes express understanding states across user utterances in a whole dialog. In the search, we used a new criterion based on efficiency in information retrieval and consistency with understanding hypotheses, which is also used to select an appropriate system response. We showed that our system can make more efficient and natural dialogs than previous ones.

AB - We regarded a dialog strategy for information retrieval as a graph search problem and proposed several novel dialog strategies that can recover from misrecognition through a spoken dialog that traverses the graph. To recover from misrecognition without seeking confirmation, our system kept multiple understanding hypotheses at each turn and searched for a globally optimal hypothesis in the graph whose nodes express understanding states across user utterances in a whole dialog. In the search, we used a new criterion based on efficiency in information retrieval and consistency with understanding hypotheses, which is also used to select an appropriate system response. We showed that our system can make more efficient and natural dialogs than previous ones.

KW - Consistency measure

KW - Efficiency measure

KW - Misrecognition

KW - Response generation

KW - Spoken dialog strategy

UR - http://www.scopus.com/inward/record.url?scp=84940321741&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84940321741&partnerID=8YFLogxK

U2 - 10.1527/tjsai.29.1

DO - 10.1527/tjsai.29.1

M3 - Article

VL - 29

SP - 1

EP - 10

JO - Transactions of the Japanese Society for Artificial Intelligence

JF - Transactions of the Japanese Society for Artificial Intelligence

SN - 1346-0714

IS - 1

ER -