Method and apparatus for providing speech recognition resolution on an application server

• US 7,016,845 B2
• Filed: 05/30/2003
• Issued: 03/21/2006
• Est. Priority Date: 11/08/2002
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Gupta, Arvind, Tuckey, Curtis, Vora, Ashish, Sprague, Kara Lynn
• Primary Examiner(s): McFadden, Susan
• Application Number:
  US10/449,174
• Publication Number:
  US 20040093216A1
• Time in Patent Office: 1,026 Days
• Field of Search: 704/251, 704/255, 704/257, 704/270.1, 704/275, 704/231, 704/270, 379/88.02, 379/88.22, 379/88.17, 379/88.16, 707/4, 706/55
• US Class Current: 704/270.1
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: H04M 3/4936 Speech interaction details speech recognition per se G10L15/00

Abstract

A method of providing speech recognition resolution on an application server in a communication network includes receiving an utterance from an end-user, for example, at a Voice Gateway in response to a prompt by an application prompt, capturing the utterance, and dispatching it via the application server to a speech recognition provider. The method further includes performing item-matching via a search algorithm, returning items matching the utterance to the application server, and returning relevant utterance matches to the application, for example, on the Voice Gateway, for communication to the user.

Claims

1. A method for providing speech recognition resolution on an application server in a communication network, comprising:

defining search spaces each comprising a dataset and identified voice searchable fields and communicating the search spaces to at least one speech recognition provider separate from the application server using an administration tool associated with the application server;

receiving an utterance from an end-user or an input application in response to a prompt by an application running on the application server;

capturing the utterance and dispatching the utterance via the application server to the speech recognition provider;

performing item-matching via the speech recognition provider, and returning items matching the utterance to the application server; and

returning relevant utterance matches to the application for communication to the user.

2. The method as in claim 1 wherein the speech recognition provider is a large-scale grammar search provider.

3. The method as in claim 2 wherein the application server runs a Web site or an Internet application, and allows the Website or the Internet application to be accessible from any browser or mobile device.

4. The method as in claim 2 wherein the large-scale grammar search provider utilizes voice search algorithms to provide an n-best search returning a weighted list of matching items.

5. The method as in claim 4 wherein the weighted list involves a confidence metric.

6. The method as in claim 2 wherein a generic search driver is utilized to communicate with any large-scale search provider, in conjunction with specific search provider drivers that format and transmit messages in a manner specific to their corresponding large-scale directory providers.

7. The method as in claim 1 wherein dispatching the utterance further comprises providing a specified search space associated with the utterance to be used by the speech recognition provider to perform the item matching.

8. The method as in claim 1 wherein a Voice Gateway, the application server and the large-scale grammar search provider are located on at least one of an Internet network and a local Internet-connected device.

9. The method as in claim 1 wherein the application server further provides a markup language page to a browser to prompt the utterance.

10. The method as in claim 9 wherein the application server communicates with other communication devices via markup language pages using at least one of VXML, XHTML, XHTML+V, SALT, HDML/WML, HTML, and MXML languages.

11. The method as in claim 1 wherein the speech recognition provider initializes and maintains search spaces in response to notification of the search spaces by the application server.

12. The method as in claim 1 wherein there are at least two different speech recognition providers and the application server utilizes the administration tool to keep track of the different speech recognition providers, and to create and manage the search spaces for each.

13. The method as in claim 1 wherein the step of capturing comprises capturing an audio recording and passing the audio recording to the large scale grammar search provider.

14. A communication system for providing speech recognition resolution on an application server in a communication network, comprising:

means for defining search spaces each comprising a dataset and identified voice searchable fields and for communicating the search spaces to at least one speech recognition provider separate from the application server using an administration tool associated with the application server;

means for receiving an utterance from an end-user or an input application in response to a prompt by an application running on the application server;

means for capturing the utterance and dispatching the utterance to the speech recognition provider;

means for performing item-matching via the speech recognition provider, and returning items matching the utterance to the application server; and

means for returning relevant utterance matches to the application for communication to the end-user.

15. The communication system as in claim 14 wherein there are at least two different speech recognition providers each of which is a large-scale grammar search engine separate from the application server and the application server monitors the different speech providers, and creates and manages the search spaces for each.

16. The communication system as in claim 15 wherein the application is accessed through a device which comprises at least a voice markup language browser.

17. The communication system as in claim 15 wherein the means for performing item-matching provides the large scale grammar search providers the ability to utilize voice search algorithms to provide an n-best search returning a weighted list of matching items.

18. A computer-readable medium having computer-executable instructions for performing a method for providing speech recognition resolution on an application server in a communication network, the method comprising:

defining search spaces each comprising a dataset and identified voice searchable fields, and communicating the search spaces to at least one speech recognition provider separate from the application server using an administration tool associated with the application server;

receiving an utterance from an end-user or an input application in respond to a prompt by an application running on the application server;

capturing the utterance and dispatching it via the application server to the speech recognition provider;

performing item-matching via the speech recognition provider, and returning items matching the utterance to the application server; and

returning relevant utterance matches to the application for communication to the user.

19. The computer-readable medium having computer-executable instructions for performing the method as in claim 18 wherein there are at least two different speech recognition providers each of which is a large scale grammar search engine separate from the application server and the application server monitors the different speech providers, and creates and manages the search spaces for each.

20. The computer-readable medium having computer-executable instructions for performing the method as in claim 19 wherein the large-scale grammar search providers utilize voice search algorithms to provide an n-best search returning a weighted list of matching items.

21. The computer-readable medium having computer-executable instructions for performing the method as in claim 19 wherein the application is accessed through a device which comprises at least a voice markup language browser.

22. The computer-readable medium having computer-executable instructions for performing the method as in claim 21 wherein the application server runs a Web site or an Internet application, and allows the Website or the application to be accessible from any browser or mobile device.