System for developing a dialog manager using modular spoken-dialog components

• US 7,421,393 B1
• Filed: 03/01/2004
• Issued: 09/02/2008
• Est. Priority Date: 03/01/2004
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: AT&T Corporation (AT&T, Inc.)
• Inventors: Di Fabbrizio, Giuseppe, Lewis, Charles Alfred
• Primary Examiner(s): Hudspeth, David R.
• Assistant Examiner(s): KOVACEK, DAVID M
• Application Number:
  US10/790,495
• Time in Patent Office: 1,646 Days
• Field of Search: 704/231, 704/246, 704251-257, 704270-278, 706/11, 706 45- 51, 706 54- 62, 707 1- 10, 379 671- 72, 379 87- 8806, 379/88.13, 379 8816- 8826, 379 9001- 9303
• US Class Current: 704/275
• CPC Class Codes: G06F 40/00 Handling natural language data speech analysis or synthesis, speech recognition G10L
• CPC Class Codes: G06F 40/191 Automatic line break hyphenation
• CPC Class Codes: G06F 40/205 Parsing
• CPC Class Codes: G06F 40/35 Discourse or dialogue representation
• CPC Class Codes: G10L 13/00 Speech synthesis; Text to speech systems
• CPC Class Codes: G10L 15/06 Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice G10L15/14 takes precedence
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/28 Constructional details of speech recognition systems
• CPC Class Codes: G10L 21/00 Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility G10L19/00 takes precedence
• CPC Class Codes: G10L 21/06 Transformation of speech into a non-audible representation, e.g. speech visualisation or speech processing for tactile aids G10L15/26 takes precedence
• CPC Class Codes: H04M 1/642 storing speech in digital form
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing

Abstract

A dialog manager and spoken dialog service having a dialog manager generated according to a method comprising selecting a top level flow controller based on application type, selecting available reusable subdialogs for each application part, developing a subdialog for each application part not having an available subdialog and testing and deploying the spoken dialog service using the selected top level flow controller, selected reusable subdialogs and developed subdialogs. The dialog manager capable of handling context shifts in a spoken dialog with a user. Application dependencies are established in the top level flow controller thus enabling the subdialogs to be reusable and to be capable of managing context shifts and mixed initiative dialogs.

Claims

1. A dialog manager for use within a spoken dialog service, the dialog manager generated according to a method comprising:

selecting a top level flow controller that is a recursive transition network (RTN) flow controller and a finite state model;

selecting available reusable subdialogs for each application part below the top level flow controller, the reusable subdialogs being isolated from application dependencies;

developing a subdialog for each application part not having an available subdialog; and

testing and deploying the spoken dialog service using the selected top level flow controller, selected reusable subdialogs and developed subdialogs, wherein the top level flow controller, reusable subdialogs and developed subdialogs interact independent of their decision model.

2. The dialog manager of claim 1, wherein the subdialogs manage mixed-initiative conversations with a user.

3. The dialog manager of claim 1, wherein application dependencies are declared outside of the subdialogs.

4. The dialog manager of claim 3, wherein the application dependencies are part of the top level flow controller.

5. The dialog manager of claim 1, wherein the available reusable subdialogs are selected from a group related to a dialog to obtain the following information:

telephone number, social security number, account number, address, and mixed initiative dialogs.

6. The dialog manager of claim 1, wherein the subdialogs provide information to the user without receiving user input.

7. The dialog manger of claim 1, wherein an available reusable subdialog is an input subdialog.

8. The dialog manager of claim 7, wherein the input subdialog further comprises a confirmation component.

9. The dialog manager of claim 7, wherein the reusable input subdialog handles silence, rejection, low confidence natural language understanding (NLU) scores and explicit information in an input dialog with a user.

10. The dialog manager of claim 9, wherein the reusable input subdialog presents at least one pre-assigned prompt to the user in special circumstances.

11. The dialog manager of claim 10, wherein the reusable input subdialog presents a confirmation prompt to the user if an NLU value is returned with a low confidence score.

12. The dialog manager of claim 6, wherein the input subdialog includes input values that may be selected from a list including:

input prompt, yes or no response, yes value name, no value name, silence category, reject category, confidence threshold and explicit confirm.

13. The dialog manager of claim 1, wherein an available reusable subdialog is a billing subdialog.

14. The dialog manager of claim 1, wherein an available reusable subdialog is a credit card subdialog.

15. The dialog manager of claim 1, wherein the available reusable subdialogs are rule-based flow controllers.

16. The dialog manager of claim 1, wherein at least one state in the RTN flow controller has a subdialog attribute that is the name of a flow controller invoked as a subdialog.

17. The dialog manager of claim 16, wherein the state in the RTN flow controller having the subdialog attribute that invokes a subdialog further comprises a set of instructions that retrieve values from a parent dialog and set values in the invoked subdialog.

18. The dialog manager of claim 17, wherein each invoked subdialog includes a set of instructions that returns control to the parent dialog and passes retrieved values from the invoked dialog to the parent dialog upon exiting the invoked subdialog.

19. The dialog manager of claim 1, further comprising implementing a local context within a dialog data file associated with the dialog manager.

20. A spoken dialog service having a dialog manager that supports context shifts in a spoken dialog, the dialog manager generated according to a method comprising:

selecting a top level dialog flow controller that is a recursive transition network (RTN) flow controller and a finite state model;

selecting available reusable subdialogs for being invoked by the top level flow controller, the reusable subdialogs being isolated from application dependencies; and

testing and deploying the spoken dialog service using the selected top level flow controller and selected reusable subdialogs, wherein when a user of the system changes the context of the spoken dialog while in a reusable subdialog, a context shift returns a context shift indication and sets a state in the top level flow controller.

21. The spoken dialog service of claim 20, wherein when a top level flow controller receives a context shift from a subdialog and invokes a new subdialog as part of processing the context shift, the new subdialog inherits information associated with the context shift.

22. The spoken dialog service of claim 21, wherein when a user of the system changes the context of the spoken dialog while in a reusable subdialog, the context shift returns a message to the parent dialog that a context shift has occurred.

23. The spoken dialog service of claim 21, wherein a context shift is triggered by user input and generates a state name where the shift goes.

24. A spoken dialog system having a dialog manager generated according to a method comprising:

selecting a recursive transition network (RTN) top level flow controller based on application type, the top level flow controller being a finite state model and having application-dependent features such that a reusable subdialog can operate independent of the decision model of the top level flow controller and the decision models of other reusable subdialogs;

determining at least one application part below the top level flow controller, each application part requiring a different flow controller;

selecting available reusable subdialogs for each application part; and

testing and deploying the spoken dialog service using the selected top level flow controller and selected reusable subdialogs.

25. The spoken dialog system of claim 24, wherein the dialog manager is further generated by developing a subdialog for each application part not having an available subdialog.

26. The spoken dialog system of claim 24, wherein an available reusable subdialog is an input subdialog.

27. The spoken dialog system of claim 26, wherein the input subdialog further comprises a confirmation component.

28. The spoken dialog system of claim 27, wherein the reusable input subdialog handles silence, rejection, low confidence natural language understanding (NLU) scores and explicit information in an input dialog with a user.

29. The spoken dialog system of claim 28, wherein the reusable input subdialog presents at least one pre-assigned prompt to the user in special circumstances.

30. The spoken dialog system of claim 27, wherein the reusable input subdialog presents a confirmation prompt to the user if an NLU value is returned with a low confidence score.

31. The spoken dialog system of claim 24, wherein the selected available reusable subdialogs are selected independent of their decision models.

32. The spoken dialog system of claim 24, wherein the selected available reusable subdialogs handle context shifts by a user by returning control and a destination state to the top level flow controller.

33. The spoken dialog system of claim 24, wherein the selected available reusable subdialogs contain no application dependencies.

34. The spoken dialog system of claim 24, wherein the selected available reusable subdialogs are selected from the group related to a dialog with the user associated with:

a telephone number, a social security number, an account number, an e-mail address and a home or business address.