TELEPHONE CALL MANAGEMENT SYSTEM AND SPEECH RECOGNITION CONTROL METHOD THEREOF

PROBLEM TO BE SOLVED: To improve satisfaction level of a user while suppressing consumption of a hardware resource when speech recognition of telephone call data is performed.SOLUTION: A speech recognition server comprises a speech recognition engine and a mode control table that holds a speech reco...

Full description

Saved in:
Bibliographic Details
Main Author ARIKAWA HIDEHISA
Format Patent
LanguageEnglish
Japanese
Published 11.01.2018
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:PROBLEM TO BE SOLVED: To improve satisfaction level of a user while suppressing consumption of a hardware resource when speech recognition of telephone call data is performed.SOLUTION: A speech recognition server comprises a speech recognition engine and a mode control table that holds a speech recognition mode for each telephone call, and the speech recognition engine comprises a mode management unit that indicates a speech recognition mode to a verbalization unit and an output analysis unit that analyzes recognition result data which is converted into text by speech recognition. The output analysis unit indicates the speech recognition mode to the mode management unit according to a result of having analyzed the recognition result data which is converted into the text by the speech recognition. The mode management unit indicates the speech recognition mode to the verbalization unit according to the indication of the output analysis unit.SELECTED DRAWING: Figure 19 【課題】通話データを音声認識する場合に、ハードウェアリソースの消費を抑えつつ、利用者の満足度を高める。【解決手段】音声認識サーバは、音声認識エンジンと、通話ごとに音声認識のモードを保持するモード制御テーブルとを有し、音声認識エンジンは、言語化部に対して、音声認識のモードを指示するモード管理部と、音声認識によりテキスト化された認識結果データを解析する出力解析部とを有する。出力解析部は、音声認識によりテキスト化された認識結果データを解析した結果にしたがって、モード管理部に音声認識のモードを指示する。モード管理部は、出力解析部の指示にしたがい言語化部に音声認識のモードを指示する。【選択図】 図19
Bibliography:Application Number: JP20160135812