Method for designing an automated speech recognition (ASR) interface for a customer call center

• US 20050261902A1
• Filed: 05/24/2004
• Published: 11/24/2005
• Est. Priority Date: 05/24/2004
• Status: Active Grant

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Sbc Knowledge Ventures, L.P.
• Inventors: Bushey, Robert R., Knott, Benjamin A., Martin, John M.
• Granted Patent: US 7,460,650 B2
• Time in Patent Office: Days
• US Class Current: 704/243
• CPC Class Codes: G10L 15/1822 Parsing for meaning understanding
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models
• CPC Class Codes: H04M 2201/40 using speech recognition speech recognition per se G10L15/00
• CPC Class Codes: H04M 2203/355 Interactive dialogue design tools, features or methods
• CPC Class Codes: H04M 3/493 Interactive information services, e.g. directory enquiries ; Arrangements therefor, e.g. interactive voice response [IVR] systems or voice portals
• CPC Class Codes: H04M 3/5232 Call distribution algorithms

Abstract

A method of designing a customer interface for a service center, such as an automated speech recognition (ASR) self-service center. Customer activity to an existing service center is monitored, providing customer model, which includes a collection of customer tasks. These tasks are assigned to action-object pairs, which are further assigned to routing destinations. Dialog modules are designed, based on the customer model data, including disambiguation dialogs.

Claims

1. A method of designing a customer interface for an automated speech recognition (ASR) call center, comprising:

collecting data for a customer model, the customer model comprising at least recorded opening statements of customers to a call center and persona data representing customer personas;

creating a task model comprising a set of action-object pairs, based on the recorded opening statements;

creating a business rules model, representing routing destinations for customer calls;

linking the customer model to the task model, such that the opening statements are assigning to action-object pairs;

linking the task model to the business rules model, such that each action-object pair is assigned to a routing destination;

creating an action-object frequency table, based on the results of the step of linking the customer model to the task model; and

designing an interface module for each action object pair;

wherein the designing step is performed in substantially the order of the frequency table, from higher frequency action-object pairs to lower frequency action-object pairs; and

wherein the designing step is based on the persona data.

2. The method of claim 1, wherein the customer interface is for a how-to-use call center application, and the step of linking the task model to the business rules model comprises linking action-object pairs to the how-to-use call center destination.

3. The method of claim 1, wherein at least one of dialog modules is activated in response to a direct routing decision.

4. The method of claim 1, wherein at least one of the dialog modules is activated in response to a clarification routing decision.

5. The method of claim 4, wherein the dialog module is operable to transfer the customer to one of two routing destinations.

6. The method of claim 4, wherein the dialog module is a disambiguation dialog module operable to present the customer with choices.

7. The method of claim 1, wherein the persona data is used to determine language alternatives.

12. The method of claim 1, wherein at least one of dialog modules is activated in response to a direct routing decision.

13. The method of claim 1, wherein at least one of the dialog modules is activated in response to a clarification routing decision.

14. The method of claim 13, wherein the dialog module is operable to transfer the customer to one of two routing destinations.

15. The method of claim 13, wherein the dialog module is a disambiguation dialog module operable to present the customer with choices.

8. A method of designing a customer interface for a self-service call center application, comprising:

recording opening statements of customers during visits to an existing service center, thereby creating a customer model representing customer tasks;

creating a task model comprising a set of action-object pairs, based on the customer model;

creating a business rules model, representing routing destinations for customer calls;

linking the customer model to the task model, such that the opening statements are assigning to action-object pairs;

linking the task model to the business rules model, such that each action-object pair is assigned to a routing destination;

creating an action-object frequency table, based on the results of the step of linking the customer model to the task model; and

designing an interface module for each action object pair;

wherein the designing step is performed in substantially the order of the frequency table, from higher frequency action-object pairs to lower frequency action-object pairs.

9. The method of claim 8, wherein the customer interface is for an ASR call center.

10. The method of claim 9, wherein the customer interface is for a how-to-use ASR call center, and the step of linking the task model to the business rules model comprises linking action-object pairs to the how-to-use call center destination.

11. The method of claim 9, wherein the customer interface is for a web-based self-service center.

16. The method of claim 8, further comprising the step of collecting customer persona data during customer visits to the existing service center.

17. The method of claim 16, wherein the persona data is used to determine language alternatives.

18. The method of claim 8, wherein the customer model data is collected during visits to a call center.

19. The method of claim 8, wherein the customer model data is collected during visits to a web site via the Internet.