Speech bookmarks in a voice user interface using a speech recognition engine and acoustically generated baseforms

• US 7,881,938 B2
• Filed: 03/27/2007
• Issued: 02/01/2011
• Est. Priority Date: 03/27/2007
• Status: Expired due to Fees

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Parker, Julia A., Muschett, Brien H.
• Primary Examiner(s): AZAD, ABUL K
• Application Number:
  US11/691,893
• Publication Number:
  US 20080243517A1
• Time in Patent Office: 1,407 Days
• Field of Search: 704/270, 704/270.1, 704/275, 704/257
• US Class Current: 704/270.1
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

A system and method for navigating a dialog hierarchy from a voice user interface (VUI) using speech bookmarks. The method can detect a user spoken command for bookmarking a location within a dialog hierarchy of a voice response system. A user spoken bookmark can be received, which is added to a personalized bookmark grammar that is associated with a user who spoke the bookmark name. A database record can be used to associate the new bookmark name with a location within the dialog hierarchy. During a subsequent interaction between the user and the voice response system, the user can speak the bookmark name, which results in a match being detected between the spoken phrase and the personalized bookmark grammar. The voice response system can then navigate to the location within bookmark hierarchy that is associated with the speech bookmark.

Claims

1. A method for navigating a dialog hierarchy from a voice user interface, the method comprising:

detecting a user-spoken command for bookmarking a location within a dialog hierarchy of a voice response system;

receiving a user-spoken bookmark name;

adding the user-spoken bookmark name to a personalized bookmark grammar associated with a user who spoke the bookmark name;

storing a record that relates the added bookmark name to the location within the dialog hierarchy; and

during a subsequent interaction between the user and the voice response system,

receiving speech input from a user, the speech input comprising a second user-spoken command, the second user-spoken command including a desired bookmark name but not an additional command identifying the desired bookmark name as a bookmark name;

in response to receiving the second user-spoken command, comparing the desired bookmark name of the second user-spoken command to the personalized bookmark grammar,

when the desired bookmark name of the second user-spoken command matches a bookmark name stored in the personalized bookmark grammar, navigating to the location within the dialog hierarchy that is related to the bookmark name.

2. The method of claim 1 further comprising:

audibly prompting for a name for a bookmark upon detecting the user spoken command, wherein the user spoken bookmark name is received as a response to the prompting step.

3. The method of claim 1 further comprising:

determining a Uniform Resource Identifier (URI) that corresponds to the location, wherein the storing step stores the determined URI in the record.

4. The method of claim 3, wherein the URI identifies a hyperlink anchor of a dialog language based document interpreted by the voice response system, wherein said dialog language is any language of a set of languages where the set of languages includes at least one language selected from a group consisting of VoiceXML and SALT.

5. The method of claim 3, wherein the URI is specific to a speech enabled application, wherein the bookmark name stored in the personalized bookmark grammar is limited to interactions involving the user and the speech enabled application, whereby other users have different user specific speech bookmarks stored within other personalized bookmark grammars.

6. The method of claim 1, wherein the personalized bookmark grammar is a voice enrolled grammar.

7. The method of claim 6, wherein the voice enrolled grammar is a Speech Recognition Grammar Specification compliant grammar.

8. The method of claim 6, wherein the voice enrolled grammar, which is used to add the speech bookmark, is configured to be referenced via a standard SRGS grammar.

9. The method of claim 1, wherein a software language in which the dialog hierarchy is specified is a dialog based language wherein said dialog language is any language of a set of languages where the set of languages includes at least one language selected from a group consisting of VoiceXML and SALT.

10. The method of claim 1, wherein the second user-spoken command consists of the desired bookmark name.

11. The method of claim 1, wherein the speech input consists of the desired bookmark name.

24. The method of claim 1, wherein determining that the user desires to access a bookmark in response to receiving the user-spoken bookmark name comprises determining that the user desired to access a bookmark based only on the bookmark name.

12. A computer system comprising:

a processor programmed to implement

a speech recognition engine to process user input regarding speech bookmarks, each speech bookmark corresponding to a dialog location of a hierarchy of menus of a speech-enabled application;

a grammar data store configured to store the user-spoken bookmark names corresponding to speech bookmarks; and

a voice user interface configured to permit voice-based interactions between a user and the speech-enabled application,

wherein the voice user interface determines that the user desires to access a bookmark in response to;

receiving speech input from a user through the voice user interface, the speech input comprising a user-spoken command, the user-spoken command including a desired bookmark name but not an additional command identifying the desired bookmark name as a bookmark name,

in response to receiving the user-spoken command, comparing the desired bookmark name to the user-spoken bookmark names, and

when the desired bookmark name matches a user-spoken bookmark name, determining that the user desires to access a bookmark corresponding to the user-spoken bookmark name.

13. The system of claim 12, wherein the hierarchy of menus is specified using a dialog based language, wherein said dialog language is any language of a set of languages where the set of languages includes at least one language selected from a group consisting of VoiceXML and SALT.

14. The system of claim 13, wherein each of the dialog locations is a URI specified location.

15. The system of claim 12, wherein the speech recognition engine is implemented using commercial off-the-shelf components.

16. The system of claim 12, wherein the speech recognition engine is implemented using open standards.

17. The system of claim 12, wherein speech recognition engine uses acoustic baseform services.

18. The system of claim 12, wherein each user of the computer system is associated with a set of user-specific speech bookmarks.

19. The system of claim 12, wherein the voice user interface is configured to present audio of the user-spoken bookmark names to the user.

20. The system of claim 19, wherein the voice user interface is configured to present the user-spoken bookmark names by audibly listing a set of user-spoken bookmark names.

21. The computer system of claim 12, wherein the user-spoken command consists of the desired bookmark name.

22. The computer system of claim 12, wherein the speech input consists of the desired bookmark name.

23. At least one storage device having encoded thereon instructions for execution on a computer system, the instructions, when executed, causing the computer system to carry out a method comprising:

detecting a user-spoken command for bookmarking a location within a dialog hierarchy of a voice response system;

receiving a user-spoken bookmark name;

adding the user-spoken bookmark name to a personalized bookmark grammar associated with a user who spoke the bookmark name;

storing a record that relates the added bookmark name to the location within the dialog hierarchy; and

during a subsequent interaction between the user and the voice response system,

receiving speech input from a user, the speech input comprising a second user-spoken command, the second user-spoken command including a desired bookmark name but not an additional command identifying the desired bookmark name as a bookmark name;

in response to receiving the second user-spoken command, comparing the desired bookmark name of the second user-spoken command to the personalized bookmark grammar,

when the desired bookmark name of the second user-spoken command matches a bookmark name stored in the personalized bookmark grammar, navigating to the location within the dialog hierarchy that is related to the bookmark name.

25. The at least one storage device of claim 23, wherein the second user-spoken command consists of the desired bookmark name.

26. The at least one storage device of claim 23, wherein the speech input consists of the desired bookmark name.