Speech recognition system for providing voice recognition services using a conversational language model

• US 20070150278A1
• Filed: 12/22/2005
• Published: 06/28/2007
• Est. Priority Date: 12/22/2005
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: International Business Machines Corporation
• Inventors: Bates, Cary, Wallenfelt, Brian
• Granted Patent: US 8,265,933 B2
• Time in Patent Office: Days
• US Class Current: 704/257
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models
• CPC Class Codes: G10L 15/197 Probabilistic grammars, e.g. word n-grams

Abstract

Embodiments of the present invention provide a method, system and article of manufacture for adjusting a language model within a voice recognition system, based on text received from an external application. The external application may supply text representing the words of one participant to a text-based conversation. In such a case, changes may be made to a language model by analyzing the external text received from the external application.

Claims

1. A computer-implemented method for providing voice recognition services using a conversational language model, comprising:

receiving, as external text, a sequence of one or more words from an external application used by a first participant to a text-based conversation;

sampling a voice pattern of a second participant to the text-based conversation;

selecting a set of n-grams that include at least some of the words from the external text, wherein each selected n-gram identifies a probability that a candidate word identified in the n-gram will follow a sequence of preceding words in the n-gram;

comparing the candidate words identified by the respective n-grams with the voice pattern to select a recognized word; and

outputting a text based representation of the recognized word.

2. The method of claim 1, wherein the external application is an instant messaging application, and wherein the external text is an instant message received by the second participant to the conversation.

3. The method of claim 2, wherein the external application is executing on a mobile telephone.

4. The method of claim 1, further comprising, adjusting the probability for a candidate word identified in at least one of the selected n-grams, after receiving the external text.

5. The method of claim 4, wherein the probability for the at least one of the selected n-grams is increased.

6. The method of claim 4, wherein the probability for the at least one of the selected n-grams is decreased.

7. The method of claim 1, further comprising, adjusting the probability for a candidate word identified in at least one of the selected n-grams, based on an identity of the first participant or second participant.

8. The method of claim 1, wherein the conversational language model maintains a history of recognized words and received external text representing the text-based conversation and maintains an identifier representing a change in speaker from the first or second participant to the other.

9. A computer-readable medium containing a program which, when executed, performs an operation for providing voice recognition services using a conversational language model, comprising:

receiving, as external text, a sequence of one or more words from an external application used by a first participant to a text-based conversation;

sampling a voice pattern of a second participant to the text-based conversation;

selecting a set of n-grams that include at least some of the words from the external text, wherein each selected n-gram identifies a probability that a candidate word identified in the n-gram will follow a sequence of preceding words in the n-gram;

comparing the candidate words identified by the respective n-grams with the voice pattern to select a recognized word; and

outputting a text based representation of the recognized word.

10. The computer-readable medium of claim 9, wherein the external application is an instant messaging application, and wherein the external text is an instant message received by the second participant to the conversation.

11. The computer-readable medium of claim 10, wherein the external application is executing on a mobile telephone.

12. The computer-readable medium of claim 9, wherein the operations further comprise, adjusting the probability for a candidate word identified in at least one of the selected n-grams, after receiving the external text.

13. The computer-readable medium of claim 12, wherein the probability for the at least one of the selected n-grams is increased.

14. The computer-readable medium of claim 12, wherein the probability for the at least one of the selected n-grams is decreased.

15. The computer-readable medium of claim 9, wherein the operations further comprise, adjusting the probability for a candidate word identified in at least one of the selected n-grams, based on an identity of the first participant or second participant.

16. The computer-readable medium of claim 9, wherein the conversational language model maintains a history of recognized words and received external text representing the text-based conversation and maintains an identifier representing a change in speaker from the first or second participant to the other.

17. A computing device comprising:

a processor; and

a memory configured to store an application that includes instructions which, when executed by the processor, cause the processor to perform operations for providing voice recognition services using a conversational language model, comprising;

receiving, as external text, a sequence of one or more words from an external application used by a first participant to a text-based conversation;

sampling a voice pattern of a second participant to the text-based conversation;

selecting a set of n-grams that include at least some of the words from the external text, wherein each selected n-gram identifies a probability that a candidate word identified in the n-gram will follow a sequence of preceding words in the n-gram;

comparing the candidate words identified by the respective n-grams with the voice pattern to select a recognized word; and

outputting a text based representation of the recognized word.

18. The computing device of claim 17;

wherein the external application is an instant messaging application, and wherein the external text is an instant message received by the second participant to the conversation.

19. The computing device of claim 17, wherein the computing device comprises a mobile telephone.

20. The computing device of claim 17, wherein the operations further comprise, adjusting the probability for a candidate word identified in at least one of the selected n-grams, after receiving the external text.

21. The computing device of claim 20, wherein the probability for the at least one of the selected n-grams is increased.

22. The computing device of claim 20, wherein the probability for the at least one of the selected n-grams is decreased.

23. The computing device of claim 17, wherein the operations further comprise, adjusting the probability for a candidate word identified in at least one of the selected n-grams, based on an identity of the first participant or second participant.

24. A computer-implemented method for providing voice recognition services using a conversational language model, comprising:

during an on-going network-based exchange of textual messages between at least a first participant and a second participant;

sampling a voice pattern of the first participant; and

converting the voice pattern into text at least in part on the basis of an incoming text message received from the second participant.