METHOD AND APPARATUS FOR PROCESSING SPOKEN SEARCH QUERIES

• US 20120259636A1
• Filed: 06/19/2012
• Published: 10/11/2012
• Est. Priority Date: 09/08/2010
• Status: Active Grant

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Sejnoha, Vladimir, Ganong, William F. III, Vozila, Paul J., Bodenstab, Nathan M., Tam, Yik-Cheung
• Granted Patent: US 8,666,963 B2
• Time in Patent Office: Days
• US Class Current: 704/235
• CPC Class Codes: G06F 16/90332 Natural language query formulation or dialogue systems

Abstract

Some embodiments relate to a method of performing a search for content on the Internet, in which a user may speak a search query and speech recognition may be performed on the spoken query to generate a text search query to be provided to a plurality of search engines. This enables a user to speak the search query rather than having to type it, and also allows the user to provide the search query only once, rather than having to provide it separately to multiple different search engines.

Claims

1. A method, implemented at least in part on a microprocessor, of performing a search for content on the Internet, the method comprising:

receiving voice input provided from a user;

generating at least one text search query for a plurality of search engines, wherein the at least one text search query is generated, at least in part, by performing speech recognition on the voice input.

2. The method of claim 1, wherein the at least one text search query comprises at least two text search queries, and wherein the act of generating further comprises:

generating a first of the at least two text search queries, at least in part by performing speech recognition on the voice input using a first language model associated with a first of the plurality of search engines; and

generating a second of the at least two text search queries, at least in part, by performing speech recognition on the voice input using a second language model, different from the first language model, associated with a second of the plurality of search engines.

3. The method of claim 2, further comprising:

issuing the first of the at least two text speech queries to the first of the plurality of search engines; and

issuing the second of the at least two text speech queries to the second of the plurality of search engines.

4. The method of claim 2, wherein the first of the at least two text search queries is different from the second of the at least two text search queries.

5. The method of claim 1, wherein the act of generating further comprises:

selecting one of a plurality of available language models to be used in performing automatic speech recognition on the voice input; and

performing automatic speech recognition on the voice input using the selected one of the plurality of language models.

6. The method of claim 5, wherein the act of selecting further comprises:

selecting the one of a plurality of available language models to be used in performing automatic speech recognition on the voice input based, at least in part, on the content of the voice input.

7. The method of claim 1, wherein the act of generating the at least one text search query further comprises performing speech recognition on the voice input using a language model, and wherein the method further comprises:

periodically updating the language model based on frequently searched terms.

8. The method of claim 1, wherein the act of generating the at least one text search query further comprises performing speech recognition on the voice input using a language model, and wherein the method further comprises:

updating the language model based on the user'"'"'s previous selection of search results.

9. At least one computer-readable medium encoded with instructions that, when executed by at least one hardware computer processor, perform a method of performing a search for content on the Internet, the method comprising:

receiving voice input provided from a user;

generating at least one text search query for a plurality of search engines, wherein the at least one text search query is generated, at least in part, by performing speech recognition on the voice input.

10. The at least one computer-readable medium of claim 9, wherein the at least one text search query comprises at least two text search queries, and wherein the act of generating further comprises:

generating a first of the at least two text search queries, at least in part by performing speech recognition on the voice input using a first language model associated with a first of the plurality of search engines; and

generating a second of the at least two text search queries, at least in part, by performing speech recognition on the voice input using a second language model, different from the first language model, associated with a second of the plurality of search engines.

11. The at least one computer-readable medium of claim 10, wherein the method further comprises:

issuing the first of the at least two text speech queries to the first of the plurality of search engines; and

issuing the second of the at least two text speech queries to the second of the plurality of search engines.

12. The at least one computer-readable medium of claim 10, wherein the first of the at least two text search queries is different from the second of the at least two text search queries.

13. The at least one computer-readable medium of claim 9, wherein the act of generating further comprises:

selecting one of a plurality of available language models to be used in performing automatic speech recognition on the voice input; and

performing automatic speech recognition on the voice input using the selected one of the plurality of language models.

14. The at least one computer-readable medium of claim 13, wherein the act of selecting further comprises:

selecting the one of a plurality of available language models to be used in performing automatic speech recognition on the voice input based, at least in part, on the content of the voice input.

15. The at least one computer-readable medium of claim 10, wherein the act of generating the at least one text search query further comprises performing speech recognition on the voice input using a language model, and wherein the method further comprises:

periodically updating the language model based on frequently searched terms.

16. The at least one computer-readable medium of claim 10, wherein the act of generating the at least one text search query further comprises performing speech recognition on the voice input using a language model, and wherein the method further comprises:

updating the language model based on the user'"'"'s previous selection of search results.

17. A system comprising:

at least one tangible memory that stores processor-executable instructions for performing a search for content via the Internet; and

at least one hardware computer processor, coupled to the at least one tangible memory, that executes the processor-executable instructions to;

receive voice input provided from a user;

generate at least one text search query for a plurality of search engines,

wherein the at least one text search query is generated, at least in part, by performing speech recognition on the voice input.

18. The system of claim 17, wherein the at least one text search query comprises at least two text search queries, and wherein the at least one hardware processor generates the at least two text search queries by:

generating a first of the at least two text search queries, at least in part by performing speech recognition on the voice input using a first language model associated with a first of the plurality of search engines; and

generating a second of the at least two text search queries, at least in part, by performing speech recognition on the voice input using a second language model, different from the first language model, associated with a second of the plurality of search engines.

19. The system of claim 18, wherein the at least one hardware computer processor executes the processor-executable instructions to:

issue the first of the at least two text speech queries to the first of the plurality of search engines; and

issue the second of the at least two text speech queries to the second of the plurality of search engines.

20. The system of claim 18, wherein the first of the at least two text search queries is different from the second of the at least two text search queries.

21. The system of claim 17, wherein the at least one hardware processor generates the at least one text search query by:

selecting one of a plurality of available language models to be used in performing automatic speech recognition on the voice input; and

performing automatic speech recognition on the voice input using the selected one of the plurality of language models.

22. The system of claim 21, wherein the at least one hardware processor selects the one of the plurality of available language models by:

selecting the one of a plurality of available language models to be used in performing automatic speech recognition on the voice input based, at least in part, on the content of the voice input.

23. The system of claim 18, wherein the at least one hardware processor generates the at least one text search query by performing speech recognition on the voice input using a language model, and wherein the at least one hardware processor executes the processor-executable instructions to:

periodically update the language model based on frequently searched terms.

24. The system of claim 18, wherein the at least one hardware processor generates the at least one text search by performing speech recognition on the voice input using a language model, and wherein the at least one hardware processor executes the processor-executable instructions to:

update the language model based on the user'"'"'s previous selection of search results.