System and method for extracting a specific situation from a conversation

• US 9,269,357 B2
• Filed: 10/09/2009
• Issued: 02/23/2016
• Est. Priority Date: 10/10/2008
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Itoh, Nobuyasu, Kurata, Gakuto, Nishimura, Masafumi
• Primary Examiner(s): JACKSON, JAKIEDA R
• Application Number:
  US12/576,540
• Publication Number:
  US 20100114575A1
• Time in Patent Office: 2,328 Days
• Field of Search: 704/251
• US Class Current: 1/1
• CPC Class Codes: G06F 16/685 using automatically derived transcript of audio data, e.g. lyrics speech recognition G10L15/00
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence

Abstract

A system, method, and computer readable article of manufacture for extracting a specific situation in a conversation. The system includes: an acquisition unit for acquiring speech voice data of speakers in the conversation; a specific expression detection unit for detecting the speech voice data of a specific expression from speech voice data of a specific speaker in the conversation; and a specific situation extraction unit for extracting, from the speech voice data of the speakers in the conversation, a portion of the speech voice data that forms a speech pattern that includes the speech voice data of the specific expression detected by the specific expression detection unit.

Claims

1. A system for extracting a specific situation in a conversation, comprising:

a processor; and

a memory with computer code instructions stored thereon, the processor and the memory, with the computer code instructions being configured to implement;

an acquisition unit for acquiring speech voice data of speakers in the conversation, the voice acquisition unit acquiring the voice data from at least one of recorded audio data and a telephone line, the speakers including a client and an agent;

a voice separation unit for identifying a specific speaker from the speech voice data of the conversation acquired by the acquisition unit, the specific speaker being one of the client and the agent;

a specific expression detection unit for detecting the speech voice data of a specific expression from speech voice data of the specific speaker in the conversation, the specific expression including at least one keyword; and

a specific situation extraction unit for automatically identifying a speech pattern that includes speech voice data of the specific expression, extracting, from the speech voice data of the speakers in the conversation, a portion of the speech voice data that includes a combination of the speech pattern and the speech voice data of the specific expression detected by the specific expression detection unit and wherein characteristics of the speech pattern in the portion meet certain conditions, the specific situation extraction unit for further determining whether a portion of the conversation that includes the speech voice data of the specific expression forms the speech pattern according to a difference in speech time between (i) the specific expression spoken by the specific speaker and (ii) the speech segment immediately before the specific expression spoken by the specific speaker and presenting the extracted portion for analysis.

2. The system according to claim 1, wherein the speech pattern is based on characteristics of the speech voice data of the speakers in the conversation and the characteristics of the speech voice data are based on speaking time of each speaker in the conversation.

3. The system according to claim 1, wherein the specific situation extraction unit extracts from the speech voice data the portion of the conversation in a determined time range with respect to the specific expression.

4. The system according to claim 1, wherein the specific situation extraction unit obtains a quantitative characteristic value representing characteristics of the portion of the conversation that forms the speech pattern in the speech voice data.

5. The system according to claim 4, wherein the specific situation extraction unit obtains, as a characteristic value of a predetermined conversation unit, a total sum of the characteristic values obtained for the portions of the conversation included in the conversation unit.

6. The system according to claim 4, wherein the specific situation extraction unit obtains the characteristic value, based on a difference in speaking time between speaking time of the specific speaker and speaking time of a different speaker, for each speaker speaking before the speech voice data of the specific expression.

7. The system according to claim 1, further comprising:

an output unit for outputting a display of the result of extracting the portion of the speech voice data that forms a speech pattern.

8. A method implemented by a computing apparatus for extracting a specific situation by analyzing a conversation, the method comprising:

acquiring, by an acquisition unit, speech voice data of the speakers in the conversation, the voice acquisition unit acquiring the voice data from at least one of recorded audio data and a telephone line, the speakers including a client and an agent;

identifying, by a voice separation unit, a specific speaker from the speech voice data of the conversation acquired by the acquisition unit, the specific speaker being one of the client and the agent;

detecting, by a specific expression detection unit, the speech voice data of a specific expression from speech voice data of the specific speaker in the conversation, the specific expression including at least one keyword;

automatically identifying a speech pattern that includes speech voice data of the specific expression by determining whether a portion of the conversation that includes the speech voice data of the specific expression forms the speech pattern according to a difference in speech time between (i) the specific expression spoken by the specific speaker and (ii) the speech segment immediately before the specific expression spoken by the specific speaker;

extracting, by a specific extraction unit, a portion of the speech voice data that includes a combination of the speech pattern and the speech voice data of the specific expression from the speech voice data of the speakers in the conversation and wherein characteristics of the speech pattern in the portion meet certain conditions, thereby extracting the specific situation by analyzing the conversation by the computing apparatus; and

presenting the extracted portion for analysis.

9. The method according to claim 8, wherein automatically identifying a speech pattern by the specific extraction unit includes:

forming the pattern based on characteristics of speech voice data of the speakers in the conversation; and

basing the characteristics of the speech voice data on speaking time of each speaker in the conversation.

10. The method according to claim 8, further including:

extracting a portion of the conversation from the speech voice data in a determined time range with respect to the specific expression.

11. The method according to claim 8, wherein extracting by the specific extraction unit includes:

obtaining a quantitative characteristic value representing characteristics of the portion of the conversation that forms the speech pattern in the speech voice data.

12. The method according to claim 11, further including:

obtaining, as a characteristic value of a predetermined conversation unit, a total sum of the characteristic values obtained for the portions of the conversation included in the conversation unit.

13. The method according to claim 11, wherein the characteristic value is based on a difference in speaking time between speaking time of the specific speaker and speaking time of a different speaker, for each speaker speaking before the speech voice data of the specific expression.

14. The method according to claim 8, further comprising:

outputting, by an output unit, a display of the result of extracting the portion of the speech voice data that forms a speech pattern.

15. A non-transitory computer readable storage device storing computer readable instructions which, when executed, cause a computer device to implement a method of extracting a specific situation by analyzing a conversation, the method comprising:

acquiring, by an acquisition unit, speech voice data of the speakers in the conversation, the acquisition unit acquiring the voice data from at least one of recorded audio data and a telephone line, the speakers including a client and an agent;

identifying, by a voice separation unit, a specific speaker from the speech voice data of the conversation acquired by the acquisition unit, the specific speaker being one of the client and the agent;

detecting, by a specific expression detection unit, the speech voice data of a specific expression from speech voice data of the specific speaker in the conversation, the specific expression including at least one keyword;

automatically identifying a speech pattern that includes the speech voice data of the specific expression by determining whether a portion of the conversation that includes the speech voice data of the specific expression forms the speech pattern according to a difference in speech time between (i) the specific expression spoken by the specific speaker and (ii) the speech segment immediately before the specific expression spoken by the specific speaker;

extracting, by a specific extraction unit, from the speech voice data of the speakers in the conversation, a portion of the speech voice data that includes a combination of the speech pattern and the speech voice data of the specific expression and wherein characteristics of the speech pattern in the portion meet certain conditions, thereby extracting the specific situation by analyzing the conversation by the computing apparatus; and

presenting the extracted portion for analysis; and

detecting, by a specific expression detection unit, the speech voice data of a specific expression from speech voice data of specific speaker in the conversation.