WEKO3
アイテム
Natural Language Dialogue Understanding on a Four-layer Plan Recognition Model
https://ipsj.ixsq.nii.ac.jp/records/59657
https://ipsj.ixsq.nii.ac.jp/records/59657f2bc6944-29b1-44fb-99ef-8c5f50918729
名前 / ファイル | ライセンス | アクション |
---|---|---|
![]() |
Copyright (c) 1992 by the Information Processing Society of Japan
|
|
オープンアクセス |
Item type | JInfP(1) | |||||||
---|---|---|---|---|---|---|---|---|
公開日 | 1992-03-31 | |||||||
タイトル | ||||||||
タイトル | Natural Language Dialogue Understanding on a Four-layer Plan Recognition Model | |||||||
タイトル | ||||||||
言語 | en | |||||||
タイトル | Natural Language Dialogue Understanding on a Four-layer Plan Recognition Model | |||||||
言語 | ||||||||
言語 | eng | |||||||
資源タイプ | ||||||||
資源タイプ識別子 | http://purl.org/coar/resource_type/c_6501 | |||||||
資源タイプ | journal article | |||||||
著者所属 | ||||||||
ATR Interpreting Telephony Research Laboratories. | ||||||||
著者所属 | ||||||||
ATR Interpreting Telephony Research Laboratories./Now with Mitsubishi Electric Corp. | ||||||||
著者所属(英) | ||||||||
en | ||||||||
ATR Interpreting Telephony Research Laboratories. | ||||||||
著者所属(英) | ||||||||
en | ||||||||
ATR Interpreting Telephony Research Laboratories./Now with Mitsubishi Electric Corp. | ||||||||
著者名 |
Hitoshi, Iida
Hidekazu, Arita
× Hitoshi, Iida Hidekazu, Arita
|
|||||||
著者名(英) |
Hitoshi, Iida
Hidekazu, Arita
× Hitoshi, Iida Hidekazu, Arita
|
|||||||
論文抄録 | ||||||||
内容記述タイプ | Other | |||||||
内容記述 | More often than written text spoken Japanese dialogue often contains fragmental utterances in which predicate substitutional expressions case element ellipsis and contracted expressions are used. In the under-standing and translation of such dialogues it is necessary to interpret each utterance in line with the developmental flow of the dialogue. This paper proposes a method for understanding a task-oriented dialogue that employs a four-layer plan recognition model. The four layers of plans are (1) interaction plans which contain knowledge that can be represented by ordered utterances between speaker and hearer for the exchange of certain informa-tion (2) communication plans which contain knowledge that allows actions for the purpose of information transfer through dialogue to be realized as a series of communicative acts (3) domain plans which contain knowledge that allows a given action to be achieved by carrying out an ordered set of acts and (4) dialogue plans which contain knowledge that allows a dialogue to be realized by means of pragmatic knowledge for dialogue development. Use of these plans makes it possible to relate each utterance to the topic of the discourse in line with the development of the dialogue and build up structures that extend over the entire dialogue. Predicate substitutions and case element ellipsis can be resolved by interpreting the fragmental utterances within the structure of the dialogue. | |||||||
論文抄録(英) | ||||||||
内容記述タイプ | Other | |||||||
内容記述 | More often than written text, spoken Japanese dialogue often contains fragmental utterances in which predicate substitutional expressions, case element ellipsis, and contracted expressions are used. In the under-standing and translation of such dialogues, it is necessary to interpret each utterance in line with the developmental flow of the dialogue. This paper proposes a method for understanding a task-oriented dialogue that employs a four-layer plan recognition model. The four layers of plans are (1) interaction plans, which contain knowledge that can be represented by ordered utterances between speaker and hearer for the exchange of certain informa-tion, (2) communication plans, which contain knowledge that allows actions for the purpose of information transfer through dialogue to be realized as a series of communicative acts, (3) domain plans, which contain knowledge that allows a given action to be achieved by carrying out an ordered set of acts, and (4) dialogue plans, which contain knowledge that allows a dialogue to be realized by means of pragmatic knowledge for dialogue development. Use of these plans makes it possible to relate each utterance to the topic of the discourse in line with the development of the dialogue, and build up structures that extend over the entire dialogue. Predicate substitutions and case element ellipsis can be resolved by interpreting the fragmental utterances within the structure of the dialogue. | |||||||
書誌レコードID | ||||||||
収録物識別子タイプ | NCID | |||||||
収録物識別子 | AA00700121 | |||||||
書誌情報 |
Journal of Information Processing 巻 15, 号 1, p. 60-71, 発行日 1992-03-31 |
|||||||
ISSN | ||||||||
収録物識別子タイプ | ISSN | |||||||
収録物識別子 | 1882-6652 | |||||||
出版者 | ||||||||
言語 | ja | |||||||
出版者 | 情報処理学会 |