DIALOGUE EXPRESSION ANALYSIS METHOD, DIALOGUE EXPRESSION ANALYSIS SYSTEM, AND DIALOGUE EXPRESSION ANALYSIS PROGRAM

To provide a dialogue expression analysis method, system and program configured to properly convert a speech content of a speaker with less vocabulary and low uniformity of expression, to improve the accuracy of retrieving a dialogue knowledge using the speech content.SOLUTION: A method includes: es...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: MORIMOTO YASUTSUGU, MAEZAWA KEI, YOKOTE KENICHI
Format: Patent
Sprache:eng ; jpn
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:To provide a dialogue expression analysis method, system and program configured to properly convert a speech content of a speaker with less vocabulary and low uniformity of expression, to improve the accuracy of retrieving a dialogue knowledge using the speech content.SOLUTION: A method includes: estimating a sub-corpus of a dialogue record on the basis of speech of first and second speakers in the dialogue record of a dialogue; extracting first unique expression used by the first speaker and second unique expression used by the second speaker from the dialogue record, for each sub-corpus; determining whether the first unique expression and the second unique expression satisfy a predetermined condition or not; and estimating, when the predetermined condition is satisfied, that the first unique expression and the second unique expression are associated with each other.SELECTED DRAWING: Figure 10 【課題】語彙や表現の一貫性が低い話者の発話内容を適切に変換して、発話内容を用いた対話知識の検索精度を向上できる対話表現分析方法、システム及びプログラムを提供する。【解決手段】方法は、対話の対話記録における第1の話者及び第2の話者の発話に基づいて対話記録のサブコーパスを推定し、サブコーパス毎に対話記録から第1の話者が用いた第1の固有表現及び第2の話者が用いた第2の固有表現を抽出し、第1の固有表現及び第2の固有表現が所定条件を充足するか否かを判定し、所定条件を充足する場合に、第1の固有表現と第2の固有表現とが対応関係を有すると推定する、各処理を有する。【選択図】図10