Intent disambiguation within a virtual agent platform

Patent metadata

Field	Value
Publication number	US-12057116-B2
Application number	US-202117162007-A
Country	US
Kind code	B2
Filing date	Jan 29, 2021
Priority date	Jan 29, 2021
Publication date	Aug 6, 2024
Grant date	Aug 6, 2024

Abstract

The present disclosure is directed techniques for executing a task or service using a virtual agent. A method includes: executing, using a virtual agent, one or more tiers of a plurality of tiers of machine learning analysis to identify a desired action to be performed based on a user command, the user command being received from an external computing device; responsive to the one or more tiers of the plurality of tiers of machine learning analysis identifying a plurality of actions associated with the user command, determining a series of inquiries to present via the external computing device, wherein each inquiry of the series of inquiries is selected based on a number of actions associated with each inquiry, and wherein each subsequent inquiry in the series of inquires is based on a user response to a preceding inquiry; identifying, based on responses to the series of inquiries, the desired action to be performed; and executing the desired action to be performed.

First claim

What is claimed is:

1. A method, comprising:
executing, using a virtual agent, one or more tiers of a plurality of tiers of machine learning analysis to identify a desired action to be performed based on a user command, the user command being received from an external computing device;
responsive to the one or more tiers of the plurality of tiers of machine learning analysis identifying a plurality of actions associated with the user command, determining a series of inquiries to present via the external computing device, wherein a first inquiry of the series of inquiries comprises a first plurality of options presented to a user, wherein a second inquiry in the series of inquires comprises a second plurality of options, wherein the first plurality of options comprise a combination of substantive options and a null option, wherein the second plurality of options comprise a new set of substantive options responsive to receiving a selection of the null option, and wherein each of the first inquiry and the second inquiry is selected based on a number of actions associated with the corresponding inquiry;
identifying, based on responses to the series of inquiries, the desired action to be performed; and
executing the desired action to be performed.

2. The method of claim 1 , wherein a number of options presented in the first and second plurality of options presented to the user is based on a current usage setting of the external computing device or a user profile.

3. The method of claim 1 , wherein a number of options presented in the first plurality of options presented to the user is between two and five options.

4. The method of claim 1 , wherein determining the series of inquiries to present via the external computing device comprises selecting an inquiry from among a plurality of inquiries that yields a fewest number of actions.

5. The method of claim 4 , wherein selecting the inquiry from among the plurality of inquiries that yields the fewest number of actions is based on a standard deviation of a number of options associated with each of the plurality of inquiries.

6. A system, comprising:
a memory; and
a processor coupled to the memory and configured to:
execute, using a virtual agent, one or more tiers of a plurality of tiers of machine learning analysis to identify a desired action to be performed based on a user command, the user command being received from an external computing device;
responsive to the one or more tiers of the plurality of tiers of machine learning analysis identifying a plurality of actions associated with the user command, determine a series of inquiries to present via the external computing device, wherein a first inquiry of the series of inquiries comprises a first plurality of options presented to a user, wherein a second inquiry in the series of inquires comprises a second plurality of options, wherein the first plurality of options comprise a combination of substantive options and a null option, wherein the second plurality of options comprise a new set of substantive options responsive to receiving a selection of the null option, and wherein each of the first inquiry is selected based on a number of actions associated with the corresponding inquiry;

identify, based on responses to the series of inquiries, the desired action to be performed; and
execute the desired action to be performed.

7. The system of claim 6 , wherein a number of options presented in the first and second plurality of options presented to the user is based on a current usage setting of the external computing device or a user profile.

8. The system of claim 6 , wherein a number of options presented in the first plurality of options presented to the user is between two and five options.

9. The system of claim 6 , wherein, to determine the series of inquiries to present via the external computing device, the processor is further configured to select an inquiry from among a plurality of inquiries that yields a fewest number of actions.

10. The system of claim 9 , wherein selecting the inquiry from among the plurality of inquiries that yields the fewest number of actions is based on a standard deviation of a number of options associated with each of the plurality of inquiries.

11. A non-transitory computer-readable device having instructions stored thereon that, when executed by at least one computing device, cause the at least one computing device to perform operations comprising:
executing, using a virtual agent, one or more tiers of a plurality of tiers of machine learning analysis to identify a desired action to be performed based on a user command, the user command being received from an external computing device;
responsive to the one or more tiers of the plurality of tiers of machine learning analysis identifying a plurality of actions associated with the user command, determining a series of inquiries to present via the external computing device, wherein a first inquiry of the series of inquiries comprises a first plurality of options presented to a user, wherein a second inquiry in the series of inquires comprises a second plurality of options, wherein the first plurality of options comprise a combination of substantive options and a null option, wherein the second plurality of options comprise a new set of substantive options responsive to receiving a selection of the null option, and wherein each of the first inquiry and the second inquiry is selected based on a number of actions associated with the corresponding inquiry;
identifying, based on responses to the series of inquiries, the desired action to be performed; and
executing the desired action to be performed.

12. The non-transitory computer-readable device of claim 11 , wherein a number of options presented in the first and second plurality of options presented to the user is based on a current usage setting of the external computing device or a user profile.

13. The non-transitory computer-readable device of claim 11 , wherein determining the series of inquiries to present via the external computing device comprises selecting an inquiry from among a plurality of inquiries that yields a fewest number of actions.

14. The non-transitory computer-readable device of claim 13 , wherein selecting the inquiry from among the plurality of inquiries that yields the fewest number of actions is based on a standard deviation of a number of options associated with each of the plurality of inquiries.

Assignees

• Salesforce Com Inc
• Salesforce Inc

Inventors

• Rodriguez Juan
• Machado Michael

Classifications

• G10L15/26

  Speech to text systems (G10L15/08 takes precedence) · CPC title

• G06F9/453

  Help systems · CPC title

• G10L15/32

  Multiple recognisers used in sequence or in parallel; Score combination systems therefor, e.g. voting systems · CPC title

• G10L2015/223

  Execution procedure of a spoken command · CPC title

• G06N20/00

  Machine learning · CPC title