Framework for language understanding systems that maximizes reuse through chained resolvers

• US 10,467,345 B2
• Filed: 08/30/2016
• Issued: 11/05/2019
• Est. Priority Date: 08/30/2016
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Radostev, Vasiliy, Sarikaya, Ruhi
• Primary Examiner(s): Riley, Marcus T
• Application Number:
  US15/252,000
• Publication Number:
  US 20180060300A1
• Time in Patent Office: 1,162 Days
• Field of Search: None
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 40/35 Discourse or dialogue representation

Abstract

Aspects herein provide third-party application authors with a resolver chaining platform that simplifies the task of creating customized resolvers to gather information from user input while also allowing those authors to chain their custom resolvers with generic resolvers provided by the platform'"'"'s host and that relate to commonly used parameter types. In particular, the present disclosure provides a method and system for authoring and using these resolver chains made up of a combination of custom and generic resolvers.

Claims

1. A system comprising:

at least one processor; and

memory encoding computer executable instructions that, when executed by the at least one processor, perform a method comprising;

receiving a selection of a first generic resolver, wherein the first generic resolver defines a first set of parameters for performing a commonly used function;

receiving a first custom resolver, wherein the first custom resolver is associated with a specific application and defines a second set of parameters for a task;

chaining the first generic resolver and the first custom resolver into a resolver chain, wherein the resolver chain is operable to extract at least one parameter from a natural language expression related to the task and pass the at least one parameter from the first generic resolver to the first custom resolver; and

storing the resolver chain for future use by a natural language understanding system.

2. The system of claim 1, wherein the resolver chain contains a sub-dialog to disambiguate the natural language expression.

3. The system of claim 1, wherein the first generic resolver resolves to a location.

4. The system of claim 3, wherein the location is selected from at least one of the following locations:

a restaurant;

a theater;

a home;

a school;

a park;

a monument;

a library;

an office;

a government building;

a store;

an address;

a place of interest; and

a cross-street.

5. The system of claim 1, wherein the first generic resolver resolves to a person.

6. The system of claim 5, wherein the person is selected from at least one of the following persons:

an artist;

a singer;

an actor;

a CEO;

an executive;

a celebrity; and

an athlete.

7. The system of claim 1, wherein the first custom resolver defines values to validate the at least one extracted parameter.

8. The system of claim 1, wherein receiving the selection of the first generic resolver comprises:

receiving a selection of the first generic resolver; and

receiving a selection of a second generic resolver, and wherein chaining the first generic resolver and the first custom resolver into the resolver chain comprises chaining the first generic resolver, the second generic resolver and the first custom resolver.

9. A method for providing parameters to a task, the method comprising:

receiving a selection of at least one generic resolver, wherein a first generic resolver of the at least one generic resolvers defines a first set of parameters for performing a commonly used function;

receiving a selection of a first custom resolver, wherein the first custom resolver is associated with a specific application and defines a second set of parameters for a task;

chaining the at least one generic resolver and the first custom resolver into a resolver chain, wherein the resolver chain is operable to extract at least one parameter from a natural language expression related to the task and pass the at least one parameter from the first generic resolver to the first custom resolver; and

storing the resolver chain for future use by a natural language understanding system.

10. The method of claim 9, wherein the resolver chain contains a sub-dialog to disambiguate the natural language expression.

11. The method of claim 9, wherein the first generic resolver resolves to a location.

12. The system of claim 11, wherein the location is selected from at least one of the following locations:

a restaurant;

a theater;

a home;

a school;

a park;

a monument;

a library;

an office;

a government building;

a store;

an address;

a place of interest; and

a cross-street.

13. The method of claim 9, wherein the first generic resolver resolves to a time interval.

14. The method of claim 13, wherein the time interval relates to an availability of a product or service.

15. The method of claim 9, wherein the first custom resolver defines values to validate the at least one extracted parameter.

16. The method of claim 9, wherein receiving the selection of at least one generic resolver comprises:

receiving a selection of the first generic resolver; and

receiving a selection of a second generic resolver, and wherein chaining the at least one generic resolver and the first custom resolver into the resolver chain comprises chaining the first generic resolver, the second generic resolver and the first custom resolver.

17. A method for using a resolver chain to extract at least one parameter from a natural language expression related to a task, the method comprising:

receiving from a user a natural language expression related to a task;

retrieving at least one resolver chain related to the task, the resolver chain comprising at least one generic resolver and at least one custom resolver; and

evaluating the natural language expression using the resolver chain, where evaluating the natural language expression comprises;

parsing the natural language expression into at least one subunit;

applying at least one generic resolver and at least one custom resolver from the resolver chain to extract at least one parameter from the at least one subunit, wherein the at least one custom resolver is associated with a specific application;

passing the at least one parameter from the at least one generic resolver to the at least one custom resolver; and

passing the at least one parameter from the at least one custom resolver to the task.

18. The method of claim 17, wherein the parameter type is at least one of the following parameter types:

a time parameter;

a location parameter;

a people parameter;

a number parameter; and

a music parameter.

19. The method of claim 17, wherein the natural language subunit is disambiguated by at least one sub-dialog.

20. The method of claim 17, wherein the parameter is validated via one or more values stored in the custom resolver.