DICTIONARY UPDATING APPARATUS, DICTIONARY UPDATING METHOD AND COMPUTER PROGRAM PRODUCT

• US 20180067920A1
• Filed: 08/29/2017
• Published: 03/08/2018
• Est. Priority Date: 09/06/2016
• Status: Active Grant

• Current Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation), Toshiba Digital Solutions Corporation (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation), Toshiba Digital Solutions Corporation (Toshiba Corporation)
• Inventors: CHO, Kenta, Goto, Kazuyuki, Miyabe, Yasunari, Shinozaki, Masahisa, Sakanushi, Keisuke, Zu, Guowei, Hirano, Kaoru
• Granted Patent: US 10,496,745 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 40/242 Dictionaries
• CPC Class Codes: G06F 40/53 Processing of non-Latin text kana-to-kanji conversion G06F40/129; vowelisation G06F40/232
• CPC Class Codes: G10L 15/04 Segmentation; Word boundary detection
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 15/065 Adaptation
• CPC Class Codes: G10L 15/187 Phonemic context, e.g. pronunciation rules, phonotactical constraints or phoneme n-grams
• CPC Class Codes: G10L 2015/0631 Creating reference templates; Clustering

Abstract

According to an embodiment, a dictionary updating apparatus includes a candidate extraction unit, a selection control unit, and a word registration unit. The candidate extraction unit extracts, based on a recognition result text obtained by a voice recognition engine performing a voice recognition processing using a word dictionary and a correction result text obtained by correcting at least a part of the recognition result text, candidates of words to be additionally registered in the word dictionary. The selection control unit generates a selection screen on which the extracted candidates are displayed selectably and, at the same time, information indicating the influence on the voice recognition processing at least when the candidates are additionally registered in the word dictionary is displayed, and accepts an operation of selecting the candidates displayed on the selection screen. The word registration unit registers additionally the candidates selected on the selection screen in the word dictionary.

Claims

1. A dictionary updating apparatus comprising:

a candidate extraction unit configured to extract, based on a recognition result text obtained by a voice recognition engine performing a voice recognition processing using a word dictionary and a correction result text obtained by correcting at least a part of the recognition result text, candidates of words to be additionally registered in the word dictionary;

a selection control unit configured to generate a selection screen on which the extracted candidates are displayed selectably and, at the same time, information indicating the influence on the voice recognition processing at least when the candidates are additionally registered in the word dictionary is displayed, and accept an operation of selecting the candidates displayed on the selection screen; and

a word registration unit configured to additionally register the candidates selected on the selection screen in the word dictionary.

2. The apparatus according to claim 1, wherein

the information indicating the influence includes information indicating an example in which the recognition result text changes before and after additionally registering the candidates in the word dictionary.

3. The apparatus according to claim 2, wherein

an example in which the candidates of the words to be additionally registered in the word dictionary is not included is preferentially displayed as the example in which the recognition result text changes.

4. The apparatus according to claim 1, wherein

the information indicating the influence includes information indicating a change in a recognition accuracy of the voice recognition engine before and after additionally registering the candidates in the word dictionary.

5. The apparatus according to claim 4, wherein

when a plurality of the candidates are displayed selectably on the selection screen, the information indicating the change in the recognition accuracy is displayed so as to change according to the selected candidates.

6. The apparatus according to claim 1, wherein

the selection control unit is configured to generate the selection screen on which a plurality of extracted candidates are displayed by arranging the candidates in descending order of a score indicating a priority of each of the candidates and, at the same time, information indicating a score increase factor or a score decrease factor of the score of each of the candidates is further displayed.

7. The apparatus according to claim 6, wherein

the information indicating the score increase factor includes information indicating that the candidate is a word matching an important word in a predetermined document.

8. The apparatus according to claim 6, wherein

the information indicating the score increase factor includes information indicating that a correction frequency of the candidate is equal to or larger than a reference value.

9. The apparatus according to claim 6, wherein

the information indicating the score increase factor includes information indicating that a recognition accuracy of the voice recognition engine is improved by additionally registering the candidates in the word dictionary.

10. The apparatus according to claim 6, wherein

the information indicating the score increase factor includes information indicating that the candidate is a word corrected by a user who is proficient in a correction operation.

11. The apparatus according to claim 6, wherein

the information indicating the score decrease factor includes information indicating that the number of characters of the candidate is a predetermined number or less.

12. A dictionary updating method comprising:

extracting, based on a recognition result text obtained by a voice recognition engine performing a voice recognition processing using a word dictionary and a correction result text obtained by correcting at least a part of the recognition result text, candidates of words to be additionally registered in the word dictionary;

generating a selection screen on which the extracted candidates are displayed selectably and, at the same time, information indicating the influence on the voice recognition processing at least when the candidates are additionally registered in the word dictionary is displayed, and accepting an operation of selecting the candidates displayed on the selection screen; and

registering additionally the candidates selected on the selection screen in the word dictionary.

13. A computer program product comprising a computer-readable medium including programmed instructions, and the instructions causing a computer to execute:

extracting, based on a recognition result text obtained by a voice recognition engine performing a voice recognition processing using a word dictionary and a correction result text obtained by correcting at least a part of the recognition result text, candidates of words to be additionally registered in the word dictionary;

generating a selection screen on which the extracted candidates are displayed selectably and, at the same time, information indicating the influence on the voice recognition processing at least when the candidates are additionally registered in the word dictionary is displayed, and accepting an operation of selecting the candidates displayed on the selection screen; and

registering additionally the candidates selected on the selection screen in the word dictionary.