System and method for maintaining a speech-recognition grammar

• US 7,542,904 B2
• Filed: 08/19/2005
• Issued: 06/02/2009
• Est. Priority Date: 08/19/2005
• Status: Expired due to Fees

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Chestnut, Kevin L., Burton, Joseph B.
• Primary Examiner(s): Hudspeth; David R
• Assistant Examiner(s): Sked; Matthew J
• Application Number:
  US11/208,366
• Publication Number:
  US 20070043566A1
• Time in Patent Office: 1,383 Days
• Field of Search: None
• US Class Current: 704/257
• CPC Class Codes: G10L 15/075 supervised, i.e. under machine guidance

Abstract

A method for distributing voice-recognition grammars includes receiving match data from a first remote element. The match data includes information associated with an attempt by the remote element to match received audio information to first stored audio data. The method also includes generating a grammar entry based on the match data. The grammar entry includes second stored audio data and a word identifier associated with the second stored audio data. Additionally, the method includes transmitting the grammar entry to a second remote element.

Claims

1. A method for distributing voice-recognition grammars, comprising:

receiving a first set of match data from a first remote element, the first set of match data comprising information associated with an attempt by the first remote element to match first received audio information to first stored audio data;

receiving a second set of match data from a second remote element, the second set of match data comprising information associated with an attempt by the second remote element to match second received audio information to the first stored audio data;

aggregating the first set of match data and the second set of match data;

deciding whether to update a grammar associated with the first stored audio data based on the aggregated match data; and

in response to deciding to update a grammar associated with the first stored audio data;

generating a grammar entry based on the aggregated match data, wherein the grammar entry includes second stored audio data and a word identifier associated with the second stored audio data;

identifying, from a plurality of stored grammars, a grammar associated with the first remote element and the second remote element;

updating the identified grammar based on the generated grammar entry; and

transmitting the updated grammar to at least one of the first remote element and the second remote element.

2. The method of claim 1, wherein the first remote element and the second remote element comprise different remote elements.

3. The method of claim 1, wherein the word identifier identifies one or more words, and wherein the second stored audio data comprises audio data associated with a mispronunciation of one or more of the identified words.

4. The method of claim 1, wherein the word identifier identifies one or more words, and wherein the second stored audio data comprises audio data associated with an abbreviation of one or more of the identified words.

5. The method of claim 1, wherein the word identifier identifies one or more names, and wherein the second stored audio data comprises audio data associated with a nickname associated with one or more of the identified names.

6. The method of claim 1, wherein transmitting the updated grammar to at least one of the first remote element and the second remote element comprises:

identifying a remote element from among a plurality of remote elements; and

transmitting the updated grammar to the identified remote element.

7. The method of claim 6, wherein at least one of the first set of match data and the second set of match data indicates a language, and wherein identifying the remote element from among the plurality of remote elements comprises identifying the remote element based on a language associated with the identified remote element.

8. The method of claim 6, wherein at least one of the first set of match data and the second set of match data indicates a location, and wherein identifying the remote element from among the plurality of remote elements comprises identifying the remote element based on a location associated with the identified remote element.

9. The method of claim 1, wherein aggregating the first set of match data and the second set of match data comprises:

selecting two or more remote elements from a plurality of remote elements; and

aggregating match data received from the selected remote elements.

10. The method of claim 1, further comprising receiving a third set of match data from the first remote element, and wherein aggregating the first set of match data and the second set of match data comprises aggregating the first set of match data, the second set of match data, and the third set of match data.

11. The method of claim 1, further comprising:

receiving, at the first remote element, audio information associated with an utterance spoken by a user;

comparing the received audio information to one or more grammar entries;

generating the first match data based on a comparison of the received audio information and the first stored audio data; and

transmitting the first match data to a grammar server.

12. The method of claim 1, further comprising identifying a plurality of remote elements associated with a first language and a second language, and wherein:

receiving the first set of match data comprises receiving the first set of match data from a first of the identified elements; and

receiving the second set of match data comprises receiving the second set of match data from a second of the identified elements.

13. The method of claim 12, wherein the first language comprises a native language associated with a subset of users of the plurality of remote elements and the second language comprises a spoken language associated with users of the plurality of remote elements.

14. A device for distributing voice-recognition grammars, comprising:

a memory operable to store processor instructions; and

a processor operable to;

receive a first set of match data from a first remote element, the first set of match data comprising information associated with an attempt by the first remote element to match first received audio information to first stored audio data;

receive a second set of match data from a second remote element, the second set of match data comprising information associated with an attempt by the second remote element to match second received audio information to the first stored audio data;

aggregate the first set of match data and the second set of match data;

decide whether to update a grammar associated with the first stored audio data based on the aggregated match data; and

in response to deciding to update a grammar associated with the first stored audio data;

generate a grammar entry based on the aggregrated match data, wherein the grammar entry includes second stored audio data and a word identifier associated with the second stored audio data;

identify, from a plurality of stored grammars, a grammar associated with the first remote element and the second remote element;

update the identified grammar based on the generated grammar entry; and

transmit the updated grammar to at least one of the first remote element and the second remote element.

15. The device of claim 14, wherein the first remote element and the second remote element comprise different remote elements.

16. The device of claim 14, wherein the word identifier identifies one or more words, and wherein the second stored audio data comprises audio data associated with a mispronunciation of one or more of the identified words.

17. The device of claim 14, wherein the word identifier identifies one or more words, and wherein the second stored audio data comprises audio data associated with an abbreviation of one or more of the identified words.

18. The device of claim 14, wherein the word identifier identifies one or more names, and wherein the second stored audio data comprises audio data associated with a nickname associated with one or more of the identified names.

19. The device of claim 14, wherein the processor is operable to transmit the updated grammar to at least one of the first remote element and the second remote element by:

identifying a remote element from among a plurality of remote elements; and

transmitting the updated grammar to the identified remote element.

20. The device of claim 19, wherein at least one of the first set of match data and the second set of match data indicates a language, and wherein the processor is operable to identify the remote element from among the plurality of remote elements by identifying the remote element based on a language associated with the identified remote element.

21. The device of claim 19, wherein at least one of the first set of match data and the second set of match data indicates a location, and wherein the processor is operable to identify the remote element from among the plurality of remote elements by identifying the remote element based on a location associated with the identified remote element.

22. The device of claim 14, wherein the processor is operable to aggregate the first set of match data and the second set of match data by:

selecting two or more remote elements from a plurality of remote elements; and

aggregating match data received from the two or more remote elements.

23. The device of claim 14, wherein the processor is further operable to receive a third set of match data from the first remote element, and wherein aggregating the first set of match data and the second set of match data comprises aggregating the first set of match data, the second set of match data, and the third set of match data.

24. The device of claim 14, wherein the processor is further operable to:

receive, at the first remote element, audio information associated with an utterance spoken by a user;

compare the received audio information to one or more grammar entries;

generate the first match data based on a comparison of the received audio information and the first stored audio data; and

transmit the first match data to a grammar server.

25. The device of claim 14, wherein the processor is further operable to identify a plurality of remote elements associated with a first language and a second language, and wherein the processor is operable to:

receive the first set of match data by receiving the first set of match data from a first of the identified elements; and

receive the second set of match data by receiving the second set of match data from a second of the identified elements.

26. The device of claim 25, wherein the first language comprises a native language associated with a subset of users of the plurality of remote elements and the second language comprises a spoken language associated with users of the plurality of remote elements.

27. A system for distributing voice-recognition grammars, comprising:

a first element operable to;

receive audio data associated with an utterance;

attempt to match the received audio information to stored audio data;

generate match information based on the attempt to match the received audio data to the stored audio data; and

transmit the match information to a server;

a server operable to;

receive a first set of match data from the first element, the first set of match data comprising information associated with an attempt by the first remote element to match first received audio information to first stored audio data;

receive a second set of match data from a second remote element, the second set of match data comprising information associated with an attempt by the second remote element to match second received audio information to the first stored audio data;

aggregate the first set of match data and the second set of match data;

decide whether to update a grammar associated with the first stored audio data based on the aggregated match data; and

in response to deciding to update a grammar associated with the first stored audio data;

generate a grammar entry based on the aggregated match data, wherein the grammar entry includes second stored audio data and a word identifier associated with the second stored audio data; and

identifying, from a plurality of stored grammars, a grammar associated with the first remote element and the second remote element;

updating the identified grammar based on the generated grammar entry;

transmit the updated grammar to at least one of the first element and the second element; and

the first element and the second element operable to receive the updated grammar from the server.

28. The system of claim 27, wherein the word identifier identifies one or more words, and wherein the second stored audio data comprises audio data associated with a mispronunciation of one or more of the identified words.

29. The system of claim 27, wherein the word identifier identifies one or more words, and wherein the second stored audio data comprises audio data associated with an abbreviation of one or more of the identified words.

30. The system of claim 27, wherein the word identifier identifies one or more names, and wherein the second stored audio data comprises audio data associated with a nickname associated with one or more of the identified names.

31. The system of claim 27, wherein the server is operable to transmit the updated grammar to at least one of the first remote element and the second remote element by:

identifying a remote element from among a plurality of remote elements; and

transmitting the updated grammar to the identified remote element.

32. The system of claim 31, wherein at least one of the first set of match data and the second set of match data indicates a language, and wherein the server is operable to identify the remote element from among the plurality of remote elements by identifying the remote element based on a language associated with the identified remote element.

33. The system of claim 31, wherein at least one of the first set of match data and the second set of match data indicates a location, and wherein the server is operable to identify the remote element from among the plurality of remote elements by identifying the remote element based on a location associated with the identified remote element.

34. The system of claim 27, wherein the server is operable to aggregate the first set of match data and the second set of match data by:

selecting two or more remote elements from a plurality of remote elements; and

aggregating match data received from the two or more remote elements.

35. The system of claim 27, wherein the server is further operable to receive receiving a third set of match data from the first remote element, and wherein aggregating the first set of match data and the second set of match data comprises aggregating the first set of match data, the second set of match data, and the third set of match data.

36. The system of claim 27, wherein the server is further operable to:

receive, at the first remote element, audio information associated with an utterance spoken by a user;

compare the received audio information to one or more grammar entries;

generate the first match data based on a comparison of the received audio information and the first stored audio data; and

transmit the first match data to a grammar server.

37. The system of claim 27, wherein the server is further operable to identify a plurality of remote elements associated with a first language and a second language, wherein the plurality of remote elements includes the first element and the second element, and wherein the server is operable to:

receive the first set of match data from the first remote element by receiving the first set of match data from a first of the identified elements; and

receive a second set of match data from the second by receiving a second set of match data from a second of the identified elements.

38. The system of claim 37, wherein the first language comprises a native language associated with a subset of users of the plurality of remote elements and the second language comprises a spoken language associated with users of the plurality of remote elements.

39. A system for distributing voice-recognition grammars, comprising:

means for receiving a first set of match data from a first remote element, the first set of match data comprising information associated with an attempt by the first remote element to match first received audio information to first stored audio data;

means for receiving a second set of match data from a second remote element, the second set of match data comprising information associated with an attempt by the second remote element to match second received audio information to the first stored audio data;

means for aggregating the first set of match data and the second set of match data;

means for deciding whether to update a grammar associated with the first stored audio data based on the aggregated match data;

means for generating a grammar entry based on the aggregated match data in response to deciding to update a grammar associated with the first stored audio data, wherein the grammar entry includes second stored audio data and a word identifier associated with the second stored audio data;

means for identifying, from a plurality of stored grammars, a grammar associated with the first remote element and the second remote element;

means for updating the identified grammar based on the generated grammar entry; and

means for transmitting the updated grammar grammar entry to at least one of the first remote element and the second remote element.

40. The system of claim 39, further comprising means for identifying a plurality of remote elements associated with a first language and a second language, and wherein:

the means for receiving the first set of match data comprises means for receiving the first set of match data from a first of the identified elements; and

the means for receiving the second set of match data comprises means for receiving the second set of match data from a second of the identified elements.

41. The system of claim 40, wherein the first language comprises a native language associated with a subset of users of the plurality of remote elements and the second language comprises a spoken language associated with users of the plurality of remote elements.