System and method for targeted tuning of a speech recognition system

• US 9,368,111 B2
• Filed: 04/25/2014
• Issued: 06/14/2016
• Est. Priority Date: 08/12/2004
• Status: Active Grant

• Current Assignee: Interactions, LLC
• Original Assignee: Interactions, LLC
• Inventors: Bushey, Robert R., Knott, Benjamin Anthony, Martin, John Mills
• Primary Examiner(s): ALBERTALLI, BRIAN LOUIS
• Application Number:
  US14/262,192
• Publication Number:
  US 20140236599A1
• Time in Patent Office: 781 Days
• US Class Current: 1/1
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 15/065 Adaptation
• CPC Class Codes: G10L 15/19 Grammatical context, e.g. disambiguation of the recognition hypotheses based on word sequence rules
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

A system and method of targeted tuning of a speech recognition system are disclosed. A particular method includes detecting that a frequency of occurrence of a particular type of utterance satisfies a threshold. The method further includes tuning a speech recognition system with respect to the particular type of utterance.

Claims

1. A method comprising:

detecting, at a computing device comprising a processor, a frequency of occurrence of a particular type of utterance;

determining, using interpretation labels previously assigned to utterances of the particular type of utterance, an accuracy value for the particular type of utterance indicating whether a speech recognition system is correctly interpreting the particular type of utterance, the accuracy value including a false alarm value for the particular type of utterance; and

determining, at the computing device, whether to tune the speech recognition system with respect to the particular type of utterance based on the frequency and based on the accuracy value for the particular type of utterance.

2. The method of claim 1, wherein the accuracy value is based on a recognition rate, a number of hits, a number of misses, or a combination thereof associated with the particular type of utterance.

3. The method of claim 1, further comprising maintaining a historical performance record.

4. The method of claim 1, wherein the particular type of utterance is an action request.

5. The method of claim 4, wherein the action request corresponds to an action object request that indicates an action and an object related to the action.

6. The method of claim 1, wherein the speech recognition system is coupled to a telephone network, a cable network, an Internet, an intranet, an extranet, or a combination thereof.

7. The method of claim 1, wherein the speech recognition system digitizes an incoming speech signal and calculates utterance parameters.

8. The method of claim 7, wherein the speech recognition system compares the utterance parameters to a library of phrases to identify a match.

9. The method of claim 1, wherein tuning the speech recognition system includes inputting a collection of utterances of the particular type of utterance into a learning module of the speech recognition system.

10. The method of claim 9, wherein inputting the collection of utterances includes playing one or more files that represent recordings of the particular type of utterance.

11. A speech recognition system, comprising:

a processor; and

a tuning engine comprising instructions that, when executed by the processor, cause the processor to perform operations including;

detecting a frequency of occurrence of a particular type of utterance;

determining, using interpretation labels previously assigned to utterances of the particular type of utterance, an accuracy value for the particular type of utterance indicating whether a speech recognition system is correctly interpreting the particular type of utterance, the accuracy value including a false alarm value for the particular type of utterance; and

determining whether to tune the speech module with respect to a particular type of utterance based on a frequency of occurrence of the particular type of utterance and based on the accuracy value for the particular type of utterance.

12. The speech recognition system of claim 11, wherein determining whether to tune the speech module includes comparing the frequency to a first threshold and comparing the accuracy value to a second threshold.

13. The speech recognition system of claim 12, further comprising a targeting engine including instructions that, when executed by the processor, cause the processor to perform operations including:

determining the frequency of occurrence of the particular type of utterance; and

determining that the frequency of occurrence of the particular type of utterance satisfies the first threshold.

14. The speech recognition system of claim 11, further comprising a repository of recorded utterances.

15. A computer readable storage device comprising instructions that, when executed by a processor, cause the processor to perform operations comprising:

detecting a frequency of occurrence of a particular type of utterance;

determining, using interpretation labels previously assigned to utterances of the particular type of utterance, an accuracy value for the particular type of utterance indicating whether a speech recognition system is correctly interpreting the particular type of utterance, the accuracy value including a false alarm value for the particular type of utterance; and

determining whether to tune the speech recognition system with respect to the particular type of utterance based on the frequency and based on the accuracy value for the particular type of utterance.