Automatic speech recognition tuning management

• US 8,694,309 B1
• Filed: 02/12/2007
• Issued: 04/08/2014
• Est. Priority Date: 02/12/2007
• Status: Active Grant

• Current Assignee: West Technology Group LLC (West Corporation)
• Original Assignee: West Corporation
• Inventors: Fisher, Aaron Scott, Pradhan, Prashanta
• Primary Examiner(s): Vo, Huyen X.
• Application Number:
  US11/673,663
• Time in Patent Office: 2,612 Days
• Field of Search: 704/231, 704/235, 704/255, 704/270, 704/270.1, 704/251, 704/257, 704 1- 10, 704/216, 704/237, 704/243, 704/244, 704/245
• US Class Current: 704/216
• CPC Class Codes: G10L 15/19 Grammatical context, e.g. disambiguation of the recognition hypotheses based on word sequence rules
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence

Abstract

A method, a computer readable medium and a system for automatic speech recognition tuning management that comprises, collecting an utterance, analyzing the utterance, correlating the collected utterance to the utterance analysis, and fetching at least one of, the collected utterance, the utterance analysis, and the correlation of the collected utterance to the utterance analysis.

Claims

1. A method for automatic speech recognition tuning management comprising:

collecting an utterance in a database;

analyzing the utterance by a processor;

providing at least one option to facilitate correct transcription of the utterance, wherein providing the at least one option comprises;

providing a transcription dictionary based upon a context provided by a translation of the utterance; and

providing a transcription spelling correction based upon the context provided by the translation of the utterance;

correlating the collected utterance to the utterance analysis;

correlating the transcription of the utterance to the translation of the utterance;

reporting a metric pertaining to the transcription of the utterance and the translation of the utterance;

providing at least two filters to create the transcription, the at least two filters include a simple filter and an advanced filter;

wherein the simple filter is configured to turn the collecting on and off;

wherein the advanced filter is configured to customize the collecting; and

fetching, by the processor, at least one of;

the collected utterance;

the utterance analysis; and

the correlation of the collected utterance to the utterance analysis.

2. The method of claim 1 comprising:

storing the collected utterance.

3. The method of claim 1 comprising:

storing the utterance analysis.

4. The method of claim 1 comprising:

receiving the fetched utterance as an audible communication.

5. The method of claim 1 comprising:

receiving the fetched analysis as a text communication.

6. The method of claim 1 comprising:

logging the time of the utterance collection.

7. The method of claim 1 comprising:

logging the date of the utterance collection.

8. The method of claim 1 comprising:

filtering the utterance collection by at least one of;

a day of week;

a time of day;

a call per hour;

a percentage of the calls; and

a number of the calls.

9. The method of claim 1 comprising:

categorizing the analyzed utterance.

10. The method of claim 1 comprising:

receiving the transcription of the collected utterance.

11. A non-transitory computer readable storage medium comprising instructions for:

receiving an utterance in a database;

providing at least one option to facilitate correct transcription of the utterance, wherein providing the at least one option comprises;

providing a transcription dictionary based upon a context provided by a translation of the utterance; and

providing a transcription spelling correction based upon the context provided by the translation of the utterance;

receiving the transcription of the utterance by a processor;

receiving the translation of the utterance by the processor;

relating the received utterance to at least one of;

the utterance transcription; and

the utterance translation;

correlating the transcription of the utterance to the translation of the utterance;

reporting a metric pertaining to the transcription of the utterance and the translation of the utterance;

providing at least two filters to create the transcription, the at least two filters include a simple filter and an advanced filter;

wherein the simple filter is configured to turn the receiving of the utterance on and off;

wherein the advanced filter is configured to customize the receiving of the utterance; and

fetching at least one of;

the received utterance;

the utterance transcription;

the utterance translation; and

the relation.

12. The non-transitory computer readable storage medium of claim 11 wherein:

the receiving of the transcription is from a plurality of users.

13. The non-transitory computer readable storage medium of claim 11 wherein:

the translation is in at least one of the following formats;

voice extensible markup language;

class; and

speech application language tags.

14. The non-transitory computer readable storage medium of claim 11 comprising instructions for:

providing a textual representation of a prompt a caller responded to while the utterance was received.

15. The non-transitory computer readable storage medium of claim 11 comprising instructions for:

sending an instant message based upon the translation to a transcriptionist.

16. The non-transitory computer readable storage medium of claim 11 comprising instructions for:

determining a confidence level of the translation based upon the transcription.

17. The non-transitory computer readable storage medium of claim 11 comprising instructions for:

determining a pathway taken by the received utterance.

18. A system for automatic speech recognition tuning management comprising:

a memory; and

a processor communicably coupled to the memory, wherein the processor;

provides at least one option to facilitate correct transcription of an utterance, wherein providing the at least one option comprises;

providing a transcription dictionary based upon a context provided by a translation of the utterance; and

providing a transcription spelling correction based upon the context provided by the translation of the utterance;

receives the transcription of the utterance;

receives the translation of the utterance;

correlates the transcription to the translation;

provide at least two filters to create the transcription, the at least two filters include a simple filter and an advanced filter;

wherein the simple filter is configured to turn the transcription of the utterance on and off;

wherein the advanced filter is configure to customize the transcription of the utterance; and

reports a metric pertaining to the transcription of the utterance and the translation of the utterance.