Numeric weighting of error recovery prompts for transfer to a human agent from an automated speech response system

• US 20070043571A1
• Filed: 08/16/2005
• Published: 02/22/2007
• Est. Priority Date: 08/16/2005
• Status: Active Grant

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: International Business Machines Corporation
• Inventors: Polkosky, Melanie, Michelini, Vanessa
• Granted Patent: US 8,073,699 B2
• Time in Patent Office: Days
• US Class Current: 704/270.100
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

A method for a speech response system to automatically transfer users to human agents. The method can establish an interactive dialog session between a user and an automated speech response system. An error score can be established when the interactive dialog session is initiated. During the interactive dialog session, responses to dialog prompts can be received. Error weights can be assigned to receive responses determined to be non-valid responses. Different non-valid responses can be assigned different error weights. For each non-valid response, the assigned error weight can be added to the error score. When a value of the error score exceeds a previously established error threshold, a user can be automatically transferred from the automated speech response system to a human agent.

Claims

1. A method used in conjunction with an automated speech response system comprising the steps of:

establishing an interactive dialog session between a user and an automated speech response system, wherein an error score is established when the interactive dialog session is initiated;

during said interactive dialog session, determining a plurality of responses to dialog prompts;

detecting whether each of said responses is a valid response;

assigning error weights to non-valid responses, wherein different non-valid responses are assigned different error weights;

for each non-valid response, adjusting said error score based upon the assigned error weight of an associated non-valid response; and

when a value of said error score exceeds a previously established error threshold, automatically transferring said user from the automated speech response system to a human agent.

2. The method of claim 1, further comprising the steps of:

when the value of the error score does not exceed the previously established error threshold, continuing the interactive dialog session between the user and the automated speech response system.

3. The method of claim 1, further comprising the steps of:

determining a severity of each of said non-valid responses, wherein the assigned error weights are based at least in part upon said determined severity.

4. The method of claim 1, said assigning step further comprising:

for each non-valid response, categorizing said non-valid response into one of a plurality of categories, wherein the error weight associated with each non-valid response score is based at least in part upon a category assigned to the non-valid response.

5. The method of claim 4, wherein the plurality of categories comprise a misrecognition category and a time out category.

6. The method of claim 4, wherein said categories comprise at least one category selected from a group consisting of an acoustic error category, a lexical error category, a syntactic error category, and a conceptual error category.

7. The method of claim 1, wherein the error weight assigned to each non-valid response is associated with a dialog prompt, said method further comprising:

when the value of the error score does not exceed the previously established threshold, the automated speech response system presenting the dialog prompt subsequent to the adjusting step.

8. The method of claim 1, wherein said dialog prompts are audible prompts including automatically generated speech, and wherein said responses to the dialogue prompts comprise speech utterances that are speech-to-text processed by the automated speech response system.

9. The method of claim 1, wherein the speech response system is an interface for a plurality of applications configured to be utilized during the interactive dialog session, wherein a value of the error score is maintained as the speech response system utilizes different ones of the plurality of applications.

10. The method of claim 1, further comprising:

during the interactive dialog session, detecting a score adjustment event; and

responsive to detecting the score adjustment event, decreasing the error score.

11. The method of claim 1, wherein said steps of claim 1 are performed by a machine in accordance with at least one computer program having a plurality of code sections that are executable by the machine.

12. The method of claim 1, wherein the steps of claim 1 are performed by at least one of a service agent and a computing device manipulated by the service agent, the steps being performed in response to a service request.

13. The method of claim 1, wherein the steps of claim 1 are performed by a machine in accordance with a plurality of machine readable instructions that are executed by the machine, and wherein a service agent responding to a service request specifically configures said machine to perform said steps.

14. A method for automatically transferring a user from a speech response system to a human agent comprising the steps of:

establishing a dialog session;

initializing an error score for the dialog session;

presenting a prompt to a user;

receiving and processing a prompt response; and

when the prompt response is not a valid response,

assigning an error weight to the prompt response;

adding the error weight to the error score;

comparing the error score to an error threshold; and

automatically transferring the user to a human agent when the error score exceeds the error threshold.

15. The method of claim 14, further comprising the steps of:

when the prompt response is a valid response, performing an appropriate programmatic action for prompt response; and

repeating said presenting and receiving steps until a method terminating event occurs, where the method terminating event comprises at least one of detecting a user initiated session termination event and detecting a system initiated transfer to the human agent.

16. The method of claim 15, further comprising the step of:

when the prompt response is not a valid response, categorizing the response into one of several categories of error responses, wherein the assigning step assigns the error weight to the prompt response based at least in part upon the category of the prompt response.

17. The method of claim 15, further comprising the step of:

when the prompt response is not a valid response, determining an error severity level to the prompt response, wherein the assigning step assigns the error weight to the prompt response based at least in part upon the determined error severity level of the prompt response.

18. The method of claim 15, wherein the steps of claim 15 are performed by at least one of a service agent and a computing device manipulated by the service agent, the steps being performed in response to a service request.

19. A speech response system comprising:

a dialog processor configured to direct interactions between users and the speech response system in accordance with a plurality of configurable dialogue prompts and in accordance with a plurality of response actions, each of the response actions being associated with different user responses to said dialog prompts; and

an error processor configured to assign numeric error weights to prompt responses that are not recognized by the speech response system as valid responses, wherein different numeric values are assigned to different prompt responses based upon at least one of a category of error associated with an invalid prompt response and a severity level of an error associated with an invalid prompt response, wherein the speech response system is configured to establish an error score for each interactive dialog session between a user and the speech response system, wherein during each interactive dialog session when invalid responses to speech response system provided prompts are detected, an numeric weight is assigned to the invalid response by the error processor and the assigned numeric weight is added to the error score, wherein the speech response system automatically transfers a user to a human agent when during an interactive dialog session the error score associated with the interactive dialog session exceeds an established error threshold.

20. A method for configuring a speech response system comprising the steps of:

identifying a service agent to configure a speech response system for a customer responsive to a service request;

the service agent configuring an error threshold for said speech response system; and

the service agent configuring a plurality of error weights, each of the error weights being associated with an error response, wherein different values are established for different ones of the error weights, wherein after the speech response system is configured by the service agent, the speech response system automatically transfers a user to a human agent when an error score for an interactive session exceeds the error threshold of the speech response system, wherein each time during the interactive session an error response is detected, the detected response is assigned one of the error weights and the assigned error weight is added to the error score.