Comment recording apparatus, method, program, and storage medium that conduct a voice recognition process on voice data

• US 8,862,473 B2
• Filed: 11/04/2010
• Issued: 10/14/2014
• Est. Priority Date: 11/06/2009
• Status: Active Grant

• Current Assignee: Ricoh Company Limited
• Original Assignee: Ricoh Company Limited
• Inventors: Iga, Soichiro, Shinnishi, Makoto, Nakatomi, Masashi
• Primary Examiner(s): He, Jialong
• Application Number:
  US12/939,646
• Publication Number:
  US 20110112835A1
• Time in Patent Office: 1,440 Days
• Field of Search: 704/235, 704/270
• US Class Current: 704/270
• CPC Class Codes: G06F 40/268 Morphological analysis
• CPC Class Codes: G10L 15/18 using natural language modelling

Abstract

A comment recording apparatus, including a voice input device and a voice output device for recording and playing back comment voice, includes a voice obtaining unit, a voice recognition unit, a morphological analysis unit, and a display generation unit. The voice obtaining unit obtains comment voice as voice data, and registers the obtained voice data to a voice database for each topic specified by a topic specification device and each comment-delivered participant identified from the voice data. The voice recognition unit conducts a voice recognition process on the voice data to obtain text information. The morphological analysis unit conducts a morphological analysis on the text information, and registers a keyword extracted from words obtained by the morphological analysis unit to a keyword database with topic and comment-delivered participant along with voice. The display generation unit displays the keyword in a matrix while relating the keyword to a topic and a comment-delivering participant.

Claims

1. A comment recording apparatus including a voice input device and a voice output device for recording and playing back comment voice, the comment recording apparatus comprising:

a voice obtaining unit to obtain comment voice, delivered by participant, as voice data using the voice input device, and to register the obtained voice data to a voice database for each topic specified by a topic specification device from among a plurality of topics registered in a topic database in advance, and each comment-delivered participant identified from the voice data;

a voice recognition unit to conduct a voice recognition process on the voice data registered in the voice database to obtain text information;

a morphological analysis unit to conduct a morphological analysis on the text information obtained by the voice recognition unit, and to register a keyword extracted from words obtained by the morphological analysis unit to a keyword database with topic and comment-delivered participant registered to the voice database along with voice; and

a display generation unit to display the keyword in a matrix while relating the keyword to a topic and a comment-delivering participant, wherein

when a current comment is delivered while the topic is changed from one topic specified by the topic specification device to another unrelated topic, the display generation unit separately displays a keyword obtained from new voice data for the another unrelated topic in a specific area in the matrix.

2. The comment recording apparatus of claim 1, wherein the topic database stores information set for each topic.

3. The comment recording apparatus of claim 1, wherein the keyword is a noun included in the text information.

4. The comment recording apparatus of claim 1, wherein the display generation unit displays a sentence consisting of the keyword displayed in the matrix and a given number of other words present before the keyword, after the keyword, or both before and after the keyword.

5. The comment recording apparatus of claim 1, wherein the display generation unit displays a sentence consisting of the keyword displayed in the matrix and a given number of other words present before the keyword, after the keyword, or both before and after the keyword, the comment recording apparatus further comprising:

a real time measuring device to measure real time of each voice; and

a voice playback unit to play back the voice data registered in the voice database by using the voice output device,

wherein the voice obtaining unit obtains the voice data with a start time of each voice based on a measurement result by the real time measuring device,

the morphological analysis unit obtains the start time of each word in the voice data, and

the voice playback unit plays back the voice data based on the start time of each word of the sentence.

6. The comment recording apparatus of claim 1,

wherein the topic database stores an explanation sentence for each topic in advance, and when the keyword displayed in the specific area matches a word in the explanation sentence for each topic obtained by conducting a morphological analysis by using the morphological analysis unit, the display generation unit displays the matched word as the keyword at a position corresponding to a corresponding topic in the matrix.

7. The comment recording apparatus of claim 1, wherein a new topic is added to the topic database based on a keyword displayed in the matrix.

8. The comment recording apparatus of claim 1, further comprising an input unit to input a level of importance and/or relatedness of keyword, displayed in the matrix, to a topic.

9. The comment recording apparatus of claim 1, wherein a keyword displayed in the matrix is delete-able, and when a keyword is deleted, the voice recognition unit re-conducts the voice recognition process.

10. A comment recording method using a comment recording apparatus for recording comment voice and playing back recorded comment voice, the comment recording apparatus including circuitry, the method comprising the steps of:

registering voice data in a voice database for each topic specified by a topic specification device from among a plurality of topics registered in a topic database in advance, and each comment-delivering participant identified from the voice data;

obtaining text information by conducting a voice recognition process on the voice data registered in the voice database by using the circuitry;

conducting a morphological analysis on the text information and extracting a keyword from words obtained by the morphological analysis;

registering the keyword in a keyword database with topic and comment-delivering participant registered to the voice database along with voice; and

displaying the keyword in a matrix by using the circuitry while relating the keyword with a topic and a comment-delivering participant, and when a current comment is delivered while the topic is changed from one topic specified by the topic specification device to another unrelated topic, separately displaying a keyword obtained from new voice data for the another unrelated topic in a specific area in the matrix.

11. A non-transitory computer-readable medium storing a program for recording comment voice using a comment recording apparatus for recording comment voice, and playing back the recorded comment voice, the comment recording apparatus including circuitry, the program comprising instructions that when executed by a computer cause the computer to execute a method of recording and playing back comment voice using the comment recording apparatus, the method comprising the steps of:

registering voice data in a voice database for each topic specified by a topic specification device from among a plurality of topics registered in a topic database in advance, and each comment-delivering participant identified from the voice data;

obtaining text information by conducting a voice recognition process on the voice data registered in the voice database by using the circuitry;

conducting a morphological analysis on the text information and extracting a keyword from words obtained by the morphological analysis;

registering the keyword in a keyword database with topic and comment-delivering participant registered to the voice database along with voice; and

displaying the keyword in a matrix by using the circuitry while relating the keyword with a topic and a comment-delivering participant, and when a current comment is delivered while the topic is changed from one topic specified by the topic specification device to another unrelated topic, separately displaying a keyword obtained from new voice data for the another unrelated topic in a specific area in the matrix.