Generating reference variations

• US 8,149,999 B1
• Filed: 12/22/2006
• Issued: 04/03/2012
• Est. Priority Date: 12/22/2006
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Tellme Networks Incorporated ([24]7.ai, Inc.)
• Inventors: Mitby, David John
• Primary Examiner(s): Sing, Simon
• Application Number:
  US11/615,063
• Time in Patent Office: 1,929 Days
• Field of Search: 379/218.01, 379/218.02, 379 8801- 8803, 455/563, 704246-256
• US Class Current: 379/88.01
• CPC Class Codes: H04M 2201/40 using speech recognition speech recognition per se G10L15/00
• CPC Class Codes: H04M 3/4936 Speech interaction details speech recognition per se G10L15/00
• CPC Class Codes: H04M 3/51 Centralised call answering arrangements requiring operator intervention , e.g. call or contact centers for telemarketing

Abstract

A method includes receiving information provided to a caller in response to an utterance from the caller. The utterance may be associated with identifying an item of interest. The method may also include generating a reference variation associated with identifying the item of interest based on the utterance.

Claims

1. A system, comprising:

a logic device configured to;

receive a telephone number provided to a caller by an operator in response to voice input from the caller,

identify a listing associated with the telephone number,

generate a variation associated with at least a portion of the listing based on the voice input, the variation comprising an alternative way to identify at least the portion of the listing;

a memory configured to store a plurality of audio files corresponding to voice inputs provided by callers and which were forwarded to operators.

2. The system of claim 1, wherein the logic device is further configured to:

at least one of generate or modify a grammar associated with performing speech recognition based on the variation.

3. The system of claim 2, wherein the logic device is further configured to:

apply the variation to a portion of a plurality of listings.

4. The system of claim 1, wherein the logic device is further configured to:

access the memory to identify a first audio file corresponding to a first voice input,

perform speech recognition on the first audio file, and

generate the variation using output from the speech recognition.

5. The system of claim 1, wherein the logic device is further configured to:

determine a number of times a first voice input was received from callers, and

modify a grammar associated with performing speech recognition after the first voice input was received from callers a predetermined number of times.

6. A method, comprising:

receiving information provided to a caller in response to an utterance from the caller, the utterance being associated with identifying an item of interest; and

generating a reference variation associated with identifying the item of interest based on the utterance;

determining a number of times a given utterance was received from callers, and

modifying a grammar associated with performing speech recognition after the first voice input has been received from callers a predetermined number of times.

7. The method of claim 6, further comprising:

generating a rule associated with performing speech recognition based on the reference variation.

8. The method of claim 7, further comprising:

using the rule for identifying a plurality of items of interest.

9. A speech recognition system, comprising:

logic configured to;

receive information provided by a human operator to a caller in response to an utterance from a caller,

identify an item of interest associated with the information provided by the operator, and

generate a variation of at least one word associated with the item of interest based on the utterance;

store audio files corresponding to voice inputs provided by callers which have been forwarded to the human operator.

10. The speech recognition system of claim 9, wherein the logic is further configured to:

generate a grammar associated with performing speech recognition based on the variation.

11. The speech recognition system of claim 9, wherein the information provided by the human operator comprises a business name.

12. The speech recognition system of claim 9, wherein the logic device is further configured to:

identify a first audio file corresponding to a first voice input,

perform speech recognition on the first audio file, and

wherein, the variation is generate from the speech recognition.

13. The speech recognition system of claim 9, wherein the logic device is further configured to:

determine a number of times a given voice input has been received from callers, and

modify a grammar associated with performing speech recognition after the given voice input was received from callers a predetermined number of times.

14. A system, comprising:

a logic device configured to;

receive a telephone number provided to a caller by an operator in response to voice input from the caller,

identify a listing associated with the telephone number, and

generate a variation associated with at least a portion of the listing based on the voice input, the variation comprising an alternative way to at least the portion of the listing;

determine a number of times a given voice input was received from callers, and

modify a grammar used for performing speech recognition after the given voice input has been received from callers a predetermined number of times.

15. The system of claim 14, wherein the logic device is further configured to:

generate a grammar associated with performing speech recognition based on the variation.

16. The system of claim 14, wherein the logic device is further configured to:

modify a grammar associated with performing speech recognition based on the variation.

17. The system of claim 14, wherein the logic device is further configured to:

apply the variation to a portion of a plurality of listings.

18. The system of claim 14, wherein, the logic device further stores a plurality of audio files corresponding to voice inputs provided by callers and which have been forwarded to operators.

19. The system of claim 14, wherein the logic device is further configured to:

identify a first audio file corresponding to a first voice input,

perform speech recognition on the first audio file, and

generate the variation using output from the speech recognition.

20. The system of claim 14, wherein, the listing provided to the caller includes a business name.