Method, apparatus, and computer-readable recording medium for improving at least one semantic unit set

• US 10,395,645 B2
• Filed: 04/22/2015
• Issued: 08/27/2019
• Est. Priority Date: 04/22/2014
• Status: Active Grant

• Current Assignee: Naver Corporation
• Original Assignee: Naver Corporation
• Inventors: Kim, Min Chul
• Primary Examiner(s): Colucci, Michael C
• Application Number:
  US14/779,037
• Publication Number:
  US 20170032778A1
• Time in Patent Office: 1,588 Days
• Field of Search: 704235, 704205, 704231, 704234, 434156, 434236, 434262, 715257, 715816
• CPC Class Codes: G10L 15/02 Feature extraction for speech recognition; Selection of recognition unit
• CPC Class Codes: G10L 15/04 Segmentation; Word boundary detection
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/025 Phonemes, fenemes or fenones being the recognition units
• CPC Class Codes: G10L 2015/0635 updating or merging of old and new templates; Mean values; Weighting
• CPC Class Codes: G10L 2015/086 Recognition of spelled words
• CPC Class Codes: G10L 25/06 the extracted parameters being correlation coefficients
• CPC Class Codes: G10L 25/51 for comparison or discrimination

Abstract

A method, system, and a computer-readable recording medium for improving a set of at least one semantic unit are provided. According to the present invention, a set of at least one semantic unit may be improved by using a phonetic sound or text.

Claims

1. A computer-implemented method for correcting errors in a captured semantic unit set including a plurality of semantic unit sets input by a user and displayed on a display, the method comprising:

receiving an analog voice signal of an improvement phonetic sound according to utterance of a user through a voice sensing unit to correct an incorrect semantic unit set within the captured semantic unit set displayed on the display;

converting, by a processor, the analog voice signal of the improvement phonetic sound to a digital signal of an improvement semantic unit set by performing at least one of a sampling process, a quantization process, and an encoding process;

specifying, by the processor, the incorrect semantic unit set to be corrected within the captured semantic unit set displayed on the display, on a basis of correlation thereof with the improvement semantic unit set converted from the analog voice signal uttered by the user; and

replacing, by the processor, the incorrect semantic unit set within the captured semantic unit set displayed on the display with the improvement semantic unit set converted from the analog voice signal uttered by the user;

wherein the incorrect semantic unit set in the captured semantic unit set displayed on the display is specified, by the processor, by measuring a length of text or a time duration of a digital phonetic signal corresponding to the improvement semantic unit set, dividing the captured semantic unit set into a plurality of portions on a basis of the length of the text or the time duration of the digital phonetic signal corresponding to the improvement semantic unit set, and comparing each of the plurality of portions of the divided captured semantic unit set with the improvement semantic unit set.

2. The method of claim 1, wherein the comparing comprises comparing features of the digital phonetic signal corresponding to each of the plurality of portions of the divided captured semantic unit set with those of the digital phonetic signal corresponding to the improvement semantic unit set.

3. The method of claim 1, wherein the comparing comprises comparing a text corresponding to each of the plurality of portions of the divided captured semantic unit set with the text corresponding to the improvement semantic unit set.

4. The method of claim 3, wherein the comparing of the texts comprises sequentially comparing the texts in terms of phonetic values or spellings.

5. The method of claim 3, wherein the comparing of the texts comprises comparing the texts as to whether the texts belong to a same category or have similar meanings.

6. The method of claim 3, wherein the comparing of the texts comprises comparing the texts on the basis of key position correlation on a keyboard.

8. A non-transitory computer-readable recording medium storing a computer program for enabling a computer to execute the method for correcting errors in the captured semantic unit set which is input by the user and including the plurality of semantic unit sets, according to claim 1.

7. A computer-implemented method for correcting errors in a semantic unit set captured from a voice according to utterance of a user and including a plurality of semantic unit sets displayed on a display, the method comprising:

receiving an improvement text inputted by the user through a user interface to correct an incorrect semantic unit set within the captured semantic unit set displayed on the display;

specifying, by a processor, an improvement semantic unit set on a basis of the improvement text inputted by the user to correct the incorrect semantic unit set;

specifying, by the processor, the incorrect semantic unit set to be corrected within the captured semantic unit set displayed on the display, on a basis of correlation thereof with the improvement semantic unit set converted from the analog voice signal uttered by the user; and

replacing, by a processor, the incorrect semantic unit set within the captured semantic unit set displayed on the display with the improvement semantic unit set converted from the analog voice signal uttered by the user;

wherein the incorrect semantic unit set in the captured semantic unit set displayed on the display is specified, by the processor, by measuring a length of text or a time duration of a digital phonetic signal corresponding to the improvement semantic unit set, dividing the captured semantic unit set into a plurality of portions on a basis of the length of the text or a the time duration of a digital phonetic signal corresponding to the improvement semantic unit set, and comparing each of the plurality of portions of the divided captured semantic unit set with the improvement semantic unit set.

9. An apparatus for correcting errors in a captured semantic unit set including a plurality of semantic unit sets input by a user and displayed on a display, the apparatus comprising:

a voice sensing unit configured to receive an analog voice signal of an improvement phonetic sound according to utterance of a user to correct an incorrect semantic unit set within the captured semantic unit set displayed on the display;

a voice processing unit configured to convert the analog voice signal of the improvement phonetic sound to a digital signal of an improvement semantic unit set by performing at least one of a sampling process, a quantization process, and an encoding process; and

a semantic unit improving unit, including at least one processor, configured to specify, the incorrect semantic unit set to be corrected within the captured semantic unit set displayed on the display, on the basis of correlation thereof with the improvement semantic unit set converted from the analog voice signal uttered by the user, and to replace the incorrect semantic unit set within the captured semantic unit set with the improvement semantic unit set displayed on the display;

wherein the incorrect semantic unit set in the captured semantic unit set displayed on the display is specified, by the at least on processor, by measuring a length of text or a time duration of a digital phonetic signal corresponding to the improvement semantic unit set, dividing the captured semantic unit set into a plurality of portions on a basis of the length of the text or the time duration of the digital phonetic signal corresponding to the improvement semantic unit set, and comparing each of the plurality of portions of the divided captured semantic unit set with the improvement semantic unit set.

10. An apparatus for correcting errors in a semantic unit set captured from a voice according to utterance of a user and including a plurality of semantic unit sets displayed on a display, the apparatus comprising:

means configured to receive an improvement text inputted by the user through a user interface to correct an incorrect semantic unit set within the captured semantic unit set displayed on the display; and

a semantic unit improving unit, including at least one processor, configured to specify an improvement semantic unit set on a basis of the improvement text inputted by the user to correct the incorrect semantic unit set, to specify, the incorrect semantic unit set to be corrected within the captured semantic unit set, on a basis of correlation thereof with the improvement semantic unit set, and to replace the incorrect semantic unit set within the captured semantic unit set with the improvement semantic unit set;

wherein the incorrect semantic unit set in the captured semantic unit set displayed on the display is specified, by the at least on processor, by measuring a length of text or the time duration of the digital phonetic signal corresponding to the improvement semantic unit set, dividing the captured semantic unit set into a plurality of portions on a basis of the length of the text or the time duration of the digital phonetic signal corresponding to the improvement semantic unit set, and comparing each of the plurality of portions of the dividing captured semantic unit set with the improvement semantic unit set.