Method and system for an automated departure strategy

• US 6,714,631 B1
• Filed: 10/31/2002
• Issued: 03/30/2004
• Est. Priority Date: 10/31/2002
• Status: Expired due to Fees

• Current Assignee: SBC Properties, L.P.
• Original Assignee: SBC Properties, L.P.
• Inventors: Chang, Hisao M., Martin, John M.
• Primary Examiner(s): Tsang, Fan
• Assistant Examiner(s): Gauthier, Gerald
• Application Number:
  US10/285,112
• Time in Patent Office: 516 Days
• Field of Search: 379/213, 379/67, 379/88.03, 379/88.01, 379/231, 379/265.01, 379/265.02, 379/88.18, 379/218.01, 379/218.02, 379/201.02, 379/88.02, 704/270, 704/275
• US Class Current: 379/88.02
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/225 Feedback of the input speech
• CPC Class Codes: H04M 3/4931 Directory assistance systems
• CPC Class Codes: H04M 3/4933 with operator assistance

Abstract

A method and system for an automated departure strategy from an automated system includes a track engine and an error engine. The track engine allows for the tracking and storing of one or more utterances spoken by a caller in response to one or more prompts. In addition, the track engine classifies each of the utterances with a confidence level. The error engine determines when one of the utterances initiates an error condition and transfers the caller to an operator on the occurrence of the error condition. In addition to transferring the caller to the operator, the error engine plays to the operator an error utterance, causing the error condition, and a preceding utterance, preceding the error utterance. Furthermore, the error engine populates an operator screen with information provided by the caller in the utterances for utterances classified with a high level of confidence.

Claims

1. A method for automated playback of caller utterances upon default to an operator in a directory assistance system, the method comprising:

receiving multiple utterances spoken by a caller interacting with a directory assistance system;

determining when one of the utterances satisfies an error condition;

transferring the caller to an operator on an occurrence of the error condition;

in response to determining that one of the utterances satisfies the error condition, automatically determining whether an automated speech recognition (ASR) engine recognized a preceding utterance among the multiple utterances with high confidence; and

if the preceding utterance has been recognized with high confidence, automatically playing the preceding utterance to the operator, and automatically playing the utterance that satisfied the error condition to the operator, such that the operator hears the utterance that satisfied the error condition in context with the preceding utterance that was recognized with high confidence.

2. A method according to claim 1, wherein the operation of determining when one of the utterances satisfies an error condition comprises:

determining that the utterance was not recognizable to the ASR engine.

3. A method according to claim 1, further comprising:

associating either a high confidence level or a lesser confidence level with each utterance that does not satisfy the error condition.

4. A method according to claim 3, wherein the operation of associating either a high confidence level or a lesser confidence level with each utterance that does not satisfy the error condition comprises:

associating a high confidence level with an utterance only if the ASR engine successfully recognized the utterance.

5. A method according to claim 1, further comprising:

in response to determining that the ASR engine recognized the preceding utterance with high confidence, automatically populating an operator screen with the preceding utterance.

6. A method according to claim 5, further comprising:

highlighting the preceding utterance in the operator screen; and

highlighting a field in the operator screen that requires additional information from the caller.

7. A method according to claim 5, further comprising:

highlighting multiple utterances from the caller in the operator screen.

8. A method according to claim 1, wherein the preceding utterance comprises a first preceding utterance, the method further comprising:

determining whether the first preceding utterance should be considered a generic utterance; and

upon a determination that the first preceding utterance should be considered a generic utterance, automatically selecting an utterance spoken prior to the first preceding utterance as the preceding utterance to be played to the operator.

9. A method according to claim 8, further comprising:

selecting the preceding utterance to be played by finding the preceding utterance that meets all of the following conditions;

spoken most immediately prior to the utterance that satisfies the error condition;

recognized with high confidence; and

should not be considered a generic utterance.

10. A method according to claim 1, further comprising:

determining whether a predetermined relationship exists between at least two preceding utterances recognized with high confidence; and

if the predetermined relationship exists between at least two preceding utterances recognized with high confidence, automatically playing the at last two preceding utterances together as the preceding utterance before playing the utterance that satisfied the error condition.

11. A method according to claim 10, wherein the operation of determining whether a predetermined relationship exists between at least two preceding utterances recognized with high confidence comprises:

determining whether the caller has successfully provided a first name and a last name.

12. A method according to claim 10, wherein the operation of determining whether a predetermined relationship exists between at least two preceding utterances recognized with high confidence comprises:

determining whether the caller has successfully provided at least two location terms.

13. A method according to claim 1, further comprising:

storing at least one of the utterances in a database.

14. A program product for automated playback of caller utterances upon default to an operator in a directory assistance system, the program product comprising:

a computer-usable medium; and

computer instructions encoded in the computer-usable medium, wherein the computer instructions, when executed, perform operation comprising;

receiving multiple utterances spoken by a caller interacting with a directory assistance system;

determining when one of the utterances satisfies an error condition;

transferring the caller to an operator on an occurrence of the error condition;

in response to determining that one of the utterances satisfies the error condition, automatically determining whether an automated speech recognition (ASR) engine recognized a preceding utterance among the multiple utterances with high confidence; and

if the preceding utterance has been recognized with high confidence, automatically playing the preceding utterance to the operator, and automatically playing the utterance that satisfied the error condition to the operator, such that the operator hears the utterance that satisfied the error condition in context with the preceding utterance that was recognized with high confidence.

15. A program product according to claim 14, wherein the operation of determining when one of the utterances satisfies an error condition comprises:

determining that the utterance was not recognizable to the ASR engine.

16. A program product according to claim 14, wherein the operations performed by the computer instructions further comprise:

associating either a high confidence level or a lesser confidence level with each utterance that does not satisfy the error condition.

17. A program product according to claim 16, wherein the operation of associating either a high confidence level or a lesser confidence level with each utterance that does not satisfy the error condition comprises:

associating a high confidence level with an utterance only if the ASR engine successfully recognized the utterance.

18. A program product according to claim 14, wherein the operations performed by the computer instructions further comprise:

in response to determining that the ASR engine recognized the preceding utterance with high confidence, automatically populating an operator screen with the preceding utterance.

19. A program product according to claim 18, wherein the operations performed by the computer instructions further comprise:

highlighting the preceding utterance in the operator screen; and

highlighting a field in the operator screen that requires additional information from the caller.

20. A program product according to claim 14, wherein the preceding utterance comprises a first preceding utterance, and wherein the operations performed by the computer instructions further comprise:

determining whether the first preceding utterance should be considered a generic utterance; and

upon a determination that the first preceding utterance should be considered a generic utterance, automatically selecting an utterance spoken prior to the first preceding utterance as the preceding utterance to be played to the operator.

21. A program product according to claim 20, wherein the operations performed by the computer instructions further comprise:

selecting the preceding utterance to be played by finding the preceding utterance that meets all of the following conditions;

spoken most immediately prior to the utterance that satisfies the error condition;

recognized with high confidence; and

should not be considered a generic utterance.

22. A program product according to claim 14, wherein the operations performed by the computer instructions further comprise:

determining whether a predetermined relationship exists between at least two preceding utterances recognized with high confidence; and

if the predetermined relationship exists between at least two preceding utterances recognized with high confidence, automatically playing the at last two preceding utterances together as the preceding utterance before playing the utterance that satisfied the error condition.

23. A program product according to claim 22, wherein the operation of determining whether a predetermined relationship exists between at least two preceding utterances recognized with high confidence comprises:

determining whether the caller has successfully provided a first name and a last name.

24. A program product according to claim 22, wherein the operation of determining whether a predetermined relationship exists between at least two preceding utterances recognized with high confidence comprises:

determining whether the caller has successfully provided at least two location terms.

25. A method for automated playback of caller utterances upon default to an operator in a directory assistance system, the method comprising:

receiving multiple utterances spoken by a caller interacting with a directory assistance system, the utterances including a most recent utterance and at least one prior utterance;

determining whether the most recent utterance satisfies an error condition;

automatically transferring the caller to an operator in response to a determination that the most recent utterance satisfies the error condition;

determining whether the prior utterance was recognized by an automated speech recognition (ASR) engine; and

in response to determining that the prior utterance was recognized by the ASR engine, automatically playing both the most recent utterance and the prior utterance to the operator, such that the operator hears an utterance that satisfied the error condition in context with a preceding utterance that was successfully recognized.

26. A method according to claim 25, wherein the operation of determining whether the most recent utterance satisfies an error condition comprises:

determining whether the ASR engine recognized the most recent utterance.

27. A method according to claim 25, further comprising:

in response to determining that the prior utterance was recognized by the ASR engine, automatically populating an operator screen with the prior utterance.

28. A method according to claim 27, further comprising:

highlighting the prior utterance in the operator screen; and

highlighting a field in the operator screen that requires additional information from the caller.

29. A method according to claim 25, wherein the utterances include the most recent utterance, a first prior utterance receive immediately before the most recent utterance, and one or more earlier utterances received before the first prior utterance, the method further comprising:

determining whether the first prior utterance should be considered generic; and

in response to determining that the first prior utterance should be considered generic, automatically selecting an utterance among the one or more earlier utterances to be played to the operator.