System and method for interactive voice services using markup language with N-best filter element

• US 7,324,942 B1
• Filed: 01/29/2002
• Issued: 01/29/2008
• Est. Priority Date: 01/29/2002
• Status: Active Grant

• Current Assignee: Microstratetegy Incorporated
• Original Assignee: Microstrategy Incorporated
• Inventors: Maass, Bodo, Mahowald, Joshua Gee-yuen
• Primary Examiner(s): Hudspeth; David
• Assistant Examiner(s): ALBERTALLI, BRIAN LOUIS
• Application Number:
  US10/058,175
• Time in Patent Office: 2,191 Days
• Field of Search: 704/270.1, 704/270, 379/88.18, 379/88.01, 379/88.16, 715/513
• US Class Current: 704/270
• CPC Class Codes: G10L 17/22 Interactive procedures; Man-machine interfaces

Abstract

A voice-enabled markup language system an N-best filter element that generates unverified possible matches to the utterance for verification to the user during an interactive voice dialog. Such an element is used when a large number of possible utterances may be made in response to an input request or similarly-sounding utterances are possible. The element in the markup language provides the structure used by a call server to dialog with the user.

Claims

1. A computer-readable storage medium comprising code programmed in a markup language for facilitating voice-enabled communication between a voice service system and an individual, the markup language comprising:

a hierarchical set of functional elements that define the capabilities of the markup language, the set of elements comprising;

a dialog element that defines a unit of interaction between the voice service system and an individual;

an input element contained in the dialog element and operative to request input from an individual during execution of a voice service; and

an n-best list filter element operative to request verification from a list of possible matches for an audibly-uttered user response;

whereby one or more of the elements are arranged to define a voice service; and

wherein the n-best filter element comprises a namespace attribute that stores results from a grammar that are confirmed as not matching the utterance.

2. The markup language of claim 1 wherein the n-best list filter element operates to cause a processing system to generate a list of possible matches for a received audible utterance.

3. The markup language of claim 1, wherein the n-best filter element comprises an expression attribute that specifies a portion on an input element to be confirmed with the user.

4. The markup language of claim 1, wherein the n-best filter element specifies a loop to go through the list of possible matches for the utterance.

5. The markup language of claim 4, wherein the n-best filter element specifies an error element to announce when a match is not found.

6. A computer-readable storage medium storing an active voice page for use in an interactive voice output comprising:

at least one dialog element contained within the container element, the dialog element comprising content for delivery to an identified user during an interactive voice broadcast;

at least one input element contained within the at least one dialog element, the at least one input element defining input to be received from the identified user during the interactive voice broadcast; and

at least one n-best list filter element operative to request verification from a list of possible matches for an audibly-uttered user response;

wherein the n-best filter element comprises a namespace attribute that stores results from a grammar that are confirmed as not matching the utterance.

7. The active voice page of claim 6 wherein the n-best list filter element operates to cause a processing system to generate a list of possible matches for a received audible utterance.

8. The active voice page of claim 6, wherein the n-best filter element comprises an expression attribute that specifies a portion on an input element to be confirmed with the user.

9. The active voice page of claim 6, wherein the n-best filter element specifies a loop to go through the list of possible matches for the utterance.

10. The active voice page of claim 9, wherein the n-best filter element specifies an error element to announce when a match is not found.

11. An interactive voice output system that dialogs with a user comprising:

a XML-based page comprising;

at least one dialog element contained within a container element, the dialog element comprising content for delivery to an identified user during an interactive voice broadcast;

at least one input element contained within the at least one dialog element, the at least one input element defining input to be received from the identified user during the interactive voice broadcast; and

at least one n-best list filter element operative to request verification from a list of possible matches for an audibly-uttered user response; and

a call server that engages a user in dialog based on the dialog element, receives input from a user and prompts the user to verify possible matches for audibly-uttered user responses that are not understood based on the XML-based page contents;

wherein the n-best filter element comprises a namespace attribute that stores results from a grammar that are confirmed as not matching the utterance.

12. The interactive voice output system of claim 11 wherein the n-best list filter element operates to cause a processing system to generate a list of possible matches for a received audible utterance.

13. The interactive voice output system of claim 11, wherein the n-best filter element comprises an expression attribute that specifies a portion on an input element to be confirmed with the user.

14. The interactive voice output system of claim 11, wherein the n-best filter element specifies a loop to go through the list of possible matches for the utterance.

15. The interactive voice output system of claim 14, wherein the n-best filter element specifies an error element to announce when a match is not found.