Using Multiple Modality Input to Feedback Context for Natural Language Understanding

• US 20130054238A1
• Filed: 08/29/2011
• Published: 02/28/2013
• Est. Priority Date: 08/29/2011
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Chambers, Robert, Kim, Michael, Bodell, Michael, Bain, John, Cross, Karen M., Gedge, Nick, Penn, Daniel Frederick, Patel, Kunal, Tecot, Edward Mark, Waltmunson, Jeremy C.
• Granted Patent: US 9,576,573 B2
• Time in Patent Office: Days
• US Class Current: 704/235
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 2015/227 of the speaker; Human-factor methodology
• CPC Class Codes: G10L 2015/228 of application context

Abstract

Input context for a statistical dialog manager may be provided. Upon receiving a spoken query from a user, the query may be categorized according to at least one context clue. The spoken query may then be converted to text according to a statistical dialog manager associated with the category of the query and a response to the spoken query may be provided to the user.

Claims

1. A method for providing input context to a speech recognizer, the method comprising:

receiving a spoken query from a user;

categorizing the spoken query according to at least one context clue;

converting the spoken query to text according to a statistical dialog manager associated with the category of the query; and

providing a response to the spoken query.

2. The method of claim 1, wherein the spoken query is associated with a web page.

3. The method of claim 2, wherein providing the response to the spoken query comprises:

performing a function associated with the web page; and

returning at least one result of performing the function.

4. The method of claim 3, further comprising providing the response as a spoken response via a text to speech conversion.

5. The method of claim 1, wherein the at least one context clue comprises at least one of the following:

a domain name, an input field name, an input field name, a profile associated with the user, a previous query, and a user-inputted value associated with a second input field.

6. The method of claim 1, wherein categorizing the spoken query comprises computing a probability that the spoken query is associated with each of a plurality of available categories.

7. The method of claim 1, wherein the query is received via a web browser.

8. The method of claim 7, wherein the statistical dialog manager executes on a server communicatively coupled to the web browser via a network.

9. The method of claim 1, wherein the query is received via a user device.

10. The method of claim 9, wherein the statistical dialog manager executes on the user device.

11. A system for providing input context to a speech recognizer, the system comprising:

a memory storage; and

a processing unit coupled to the memory storage, wherein the processing unit is operable to;

receive a plurality of user data,

identify a plurality of categories associated with the plurality of user data,

build a plurality of statistical dialog managers, wherein each of the plurality of statistical dialog managers is associated with one of the plurality of categories,

receive a spoken user query,

categorize the spoken user query, and

provide a response to the user query via the statistical dialog manager associated with the category of the user query.

12. The system of claim 11, wherein the plurality of user data comprises a plurality of messages exchanged between users.

13. The system of claim 11, wherein the plurality of user data comprises a plurality of session logs.

14. The system of claim 13, wherein the plurality of session logs are associated with a web page comprising an interactive form.

15. The system of claim 11, wherein the plurality of user data is associated with a single user.

16. The system of claim 11, wherein the plurality of user data is aggregated from a plurality of user.

17. The system of claim 16, wherein the plurality of aggregated user data is anonymized.

18. The system of claim 11 wherein the processing unit is operative to categorize the user query according to at least one of the following:

a domain name, an input field name, an input field name, a profile associated with the user, a previous query, and a user-inputted value associated with a second input field.

19. The system of claim 11, wherein being operative to provide a response to the user query via the statistical dialog manager associated with the category of the user query comprises being operative to:

convert the categorized spoken query to text via the statistical dialog manager associated with the category of the spoken query; and

apply the converted query to a search engine.

20. A computer-readable medium which stores a set of instructions which when executed performs a method for providing input context to a speech recognizer, the method executed by the set of instructions comprising:

collecting a plurality of user data associated with a web page, wherein the plurality of user data comprises a plurality of inputs and responses associated with an interactive form of the web page;

determining whether at least one element of the plurality of user data comprises a private data element;

in response to determining that the at least one element of the plurality of user data comprises the private data element, discarding the private data element;

identifying a plurality of categories associated with the plurality of user data;

building a plurality of statistical dialog managers, wherein each of the plurality of statistical dialog managers is associated with one of the plurality of categories;

receiving a spoken user query via a web browser application in communication with a web page;

categorizing the spoken user query according to at least one context clue provided by the web browser, wherein the at least one context clue comprises at least one of the following;

a domain name, an input field name, an input field name, a profile associated with the user, a previous query, and a user-inputted value associated with a second input field;

converting the categorized spoken user query to text via the statistical dialog manager associated with the category of the spoken user query;

applying the converted query to a search engine; and

providing a response to the spoken user query according the statistical dialog manager associated with the category of the user query via the web browser application.