METHOD AND SYSTEM HAVING HYPOTHESIS TYPE VARIABLE THRESHOLDS

• US 20080228486A1
• Filed: 03/13/2007
• Published: 09/18/2008
• Est. Priority Date: 03/13/2007
• Status: Active Grant

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: International Business Machines Corporation
• Inventors: Pethachi, Pichappan, Goel, Vaibhava, Gopinath, Ramesh, Claiborn, David
• Granted Patent: US 8,725,512 B2
• Time in Patent Office: Days
• US Class Current: 704/257
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

A method (and system) for spoken dialog confirmation classifies a plurality of spoken dialog hypotheses, and assigns a threshold to each class of spoken dialog hypotheses.

Claims

1. A method of spoken dialog confirmation, comprising:

classifying a plurality of spoken dialog hypotheses into a plurality of classes of spoken dialog hypotheses; and

assigning a threshold to each of said classes of spoken dialog hypotheses.

2. The method of claim 1, further comprising:

receiving an utterance;

generating a spoken dialog hypothesis and a confidence value for said utterance;

determining a corresponding threshold for said spoken dialog hypothesis based upon said classifying; and

comparing said confidence value of said spoken dialog hypothesis to each threshold.

3. The method of claim 2, further comprising accepting said spoken dialog hypothesis if said confidence value exceeds said threshold.

4. The method of claim 2, further comprising rejecting said spoken dialog hypothesis if said confidence value does not exceed said threshold.

5. The method of claim 1, wherein said spoken dialog hypotheses are classified according to a frequency of occurrence of said spoken dialog hypotheses.

6. The method of claim 1, wherein said assigning a threshold to each class comprises assigning a confirmation and rejection threshold to each class, and further comprising:

receiving an utterance;

generating a spoken dialog hypothesis and a confidence value for said utterance;

determining corresponding rejection and confirmation thresholds for said spoken dialog hypothesis based upon said classifying;

comparing said confidence value of said spoken dialog hypothesis to said thresholds;

confirming said spoken dialog hypothesis if said confidence value does not exceed said confirmation threshold but it does exceed the said rejection threshold;

accepting said spoken dialog hypothesis if said confidence value exceeds said confirmation threshold;

confirming said spoken dialog hypothesis if said confidence value does not exceed said confirmation threshold but exceeds said rejection threshold; and

rejecting said spoken dialog hypothesis if said confidence value does not exceed said rejection threshold and said spoken dialog hypothesis is not confirmed.

7. The method of claim 1, wherein said spoken dialog hypotheses are classified according to a difficulty of recognition.

8. A program embodied in a computer readable medium executable by a digital processing unit comprising instructions for performing the method of claim 1.

18. A program embodied in a computer readable medium executable by a digital processing system for spoken dialog confirmation, said program comprising instructions for executing the method of claim 1.

9. A system for spoken dialog confirmation, comprising:

a spoken dialog hypothesis classifier that classifies a plurality of hypotheses into a plurality of classes of spoken dialog hypotheses; and

a threshold assigner that assigns a threshold to each of said classes of spoken dialog hypotheses.

10. The system of claim 9, further comprising a spoken dialog hypothesis generator that generates a spoken dialog hypothesis and a confidence value for an input utterance.

11. The system of claim 10, further comprising a comparator that compares said confidence value to said threshold.

12. The system of claim 11, wherein said comparator outputs said spoken dialog hypothesis as an accepted spoken dialog hypothesis if said confidence value exceeds said threshold.

13. The system of claim 11, wherein said comparator outputs said spoken dialog hypothesis as a rejected spoken dialog hypothesis if said confidence value does not exceed said threshold.

14. The system of claim 9, wherein said threshold assigner assigns said threshold based upon the classification by said spoken dialog hypothesis classifier.

15. The system of claim 9, wherein said spoken dialog hypothesis classifier classifies said plurality of spoken dialog hypotheses based upon a frequency of occurrence of said spoken dialog hypotheses.

16. The system of claim 15, wherein said threshold is smaller as the frequency of occurrence is larger.

17. The system of claim 15, wherein said threshold is larger as the frequency of occurrence is smaller.

19. A system for spoken dialog confirmation, comprising:

means for classifying a plurality of spoken dialog hypotheses in a plurality of classes of spoken dialog hypotheses; and

means for assigning a threshold to each of said classes of spoken dialog hypotheses.

20. The system of claim 19, further comprising:

means for receiving an utterance;

means for generating a spoken dialog hypothesis and a confidence value for said utterance;

means for determining a corresponding threshold for said spoken dialog hypothesis based upon said classifying; and

means for comparing said confidence value of said spoken dialog hypothesis to said threshold.