Using a natural language model to interface with a closed domain system

Patent metadata

Field	Value
Publication number	US-12057113-B2
Application number	US-202318329839-A
Country	US
Kind code	B2
Filing date	Jun 6, 2023
Priority date	Mar 31, 2021
Publication date	Aug 6, 2024
Grant date	Aug 6, 2024

Abstract

In various examples, systems and methods of the present disclosure combine open and closed dialog systems into an intelligent dialog management system. A text query may be processed by a natural language understanding model trained to associate the text query with a domain tag, intent classification, and/or input slots. Using the domain tag, the natural language understanding model may identify information in the text query corresponding to input slots needed for answering the text query. The text query and related information may then be passed to a dialog manager to direct the text query to the proper domain dialog system. Responses retrieved from the domain dialog system may be provided to the user via text output and/or via a text to speech component of the dialog management system.

First claim

What is claimed is:

1. A method of processing input comprising:
receiving an input query to an open domain query system;
associating, using a natural language understanding model, the input query with one or more closed domain query systems;
sending at least a portion of the input query to the one or more closed domain query systems;
receiving one or more responses from the one or more closed domain query systems; and
processing, using the open domain query system, the one or more responses from the one or more closed domain query systems to obtain a response to the input query.

2. The method of claim 1 , wherein the associating the input query with the one or more closed domain query systems is performed using the open domain query system.

3. The method of claim 2 , wherein the open domain query system modifies the input query to match an expected format for the one or more closed domain query systems prior to the sending of the at least the portion of the input query to the one or more closed domain query systems.

4. The method of claim 1 , further comprising:
determining, using the natural language understanding model, that the response to the input query is not associated with any of the one or more closed domain query systems; and
responsive to the determining, causing an output of the response to the input query.

5. The method of claim 1 , further comprising:
associating, using a natural language model, the response to the input query with the one or more closed domain query systems;
sending at least a portion of the response to the one or more closed domain query systems;
receiving one or more further responses from the one or more closed domain query systems;
processing, using the open domain query system, the one or more further responses from the one or more closed domain query systems; and
processing, using the open domain query system, the one or more further responses from the one or more closed domain query systems to obtain an updated response to the input query.

6. The method of claim 5 , further comprising:
determining, using the natural language understanding model, that the updated response to the input query is not associated with any of the one or more closed domain query systems; and
responsive to the determining, causing an output of the updated response.

7. The method of claim 1 , wherein the sending at least the portion of the input query to the one or more closed domain query systems is performed using one or more plug-ins or one or more application programming interfaces (APIs) associated with the one or more closed domain query systems.

8. A system comprising:
one or more processing units to perform operations comprising:
processing a first query submitted to a dialog management system to determine at least one domain corresponding to the first query;
identifying at least one closed domain dialog system corresponding to the at least one domain;
generating a second query corresponding to the at least one closed domain dialog system;
applying the second query to the at least one closed domain dialog system to cause a generation of a closed domain response to the second query using the at least one closed domain system; and
providing, based at least on the closed domain response, an open domain response to the first query.

9. The system of claim 8 , wherein the first query comprises a text input and the processing the first query comprises performing natural language processing on at least a portion of the text input to determine the at least one domain corresponding to the first query.

10. The system of claim 8 , wherein the at least one closed domain dialog system is associated with one or more policies, and the generating the second query comprises generating the second query based at least on the one or more policies.

11. The system of claim 10 , wherein the one or more policies include at least one policy for one or more input slots for an interface corresponding to the at least one closed domain dialog system.

12. The system of claim 11 , wherein the generating the second query comprises identifying information from the first query to fill the one or more input slots.

13. The system of claim 12 , wherein the interface comprises a plug-in.

14. The system of claim 8 , wherein the open domain response is further provided based at least on the first query.

15. One or more processing units comprising processing circuitry to perform operations comprising:
determining, based at least on analyzing an input query using a natural language model of an open domain dialog system, that the input query corresponds to a closed domain dialog system;
generating a request for the closed domain dialog system using query data corresponding to the input query; and
generating, using the open domain dialog system, a response to the input query based at least on responsive data received from the closed domain dialog system in response to the request.

16. The one or more processing units of claim 15 , wherein the determining comprises evaluating the query data in view of at least one of one or more rules or one or more policies corresponding to the closed domain dialog system.

17. The one or more processing units of claim 15 , wherein the generating the request for the closed domain dialog system is based at least on at least one of one or more rules or one or more policies corresponding to the closed domain dialog system.

18. The one or more processing units of claim 15 , wherein the operations further comprise routing, using one or more plug-ins corresponding to the closed domain dialog system, the query data to the closed domain dialog system.

19. The one or more processing units of claim 15 , wherein the input query includes a textual query, and the analyzing includes processing the textual query using the natural language model.

20. The one or more processing units of claim 15 , wherein the request is sent to the closed domain dialog system using one or more application programming interfaces (APIs) or one or more plug-ins.

Assignees

• Nvidia Corp

Inventors

• Das Shubhadeep
• Bhattacharya Sumit
• Kumar Ratin

Classifications

• G10L15/22

  Procedures used during a speech recognition process, e.g. man-machine dialogue · CPC title

• G10L15/30

  Distributed recognition, e.g. in client-server systems, for mobile phones or network applications · CPC title

• G10L13/02

  Methods for producing synthetic speech; Speech synthesisers · CPC title

• G06F16/3329 Primary

  Natural language query formulation · CPC title

• G10L2015/223

  Execution procedure of a spoken command · CPC title