Conversational agent

• US 9,971,766 B2
• Filed: 02/17/2017
• Issued: 05/15/2018
• Est. Priority Date: 04/20/2012
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Maluuba Inc. (Microsoft Corporation)
• Inventors: Pasupalak, Sam, Pantony, Joshua R., Hsu, Wilson, Wu, Zhiyuan, Tregenza, Phil, Suleman, Kaheer, Simpson, James, McNamara, Andrew, Ismail, Tareq
• Primary Examiner(s): SINGH, SATWANT K
• Application Number:
  US15/436,429
• Publication Number:
  US 20170228367A1
• Time in Patent Office: 452 Days
• Field of Search: None
• CPC Class Codes: G06F 16/90332 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 40/20 Natural language analysis semantic analysis of natural language G06F40/30
• CPC Class Codes: G06F 40/295 Named entity recognition
• CPC Class Codes: G06F 40/35 Discourse or dialogue representation
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: H04M 3/4936 Speech interaction details speech recognition per se G10L15/00

Abstract

A method, system, and computer program product provide a conversation agent to process natural language queries expressed by a user and perform commands according to the derived intention of the user. A natural language processing (NLP) engine derives intent using conditional random fields to identify a domain and at least one task embodied in the query. The NLP may further identify one or more subdomains, and one or more entities related to the identified command. A template system creates a data structure for information relevant to the derived intent and passes a template to a services manager for interfacing with one or more services capable of accomplishing the task. A dialog manager may elicit more entities from the user if required by the services manager and otherwise engage in conversation with the user. In one embodiment, the conversational agent allows a user to engage in multiple conversations simultaneously.

Claims

1. A computing device comprising one or more processors and one or more non-transitory storage devices storing instructions that, when executed by the one or more processors, configure the computing device to:

receive at least one user query from an input device, wherein the at least one user query includes a user intent and at least one command, the at least one user intent being associated with a set of functions to be carried out by the computing device;

derive, using a natural language processing (NLP) engine, the user intent from the user query;

identify, using the NLP engine, at least one domain and the at least one command associated with the derived user intent, wherein the NLP engine performs entity extraction on the at least one user query to identify at least one entity associated with the at least one command, wherein the at least one command is associated with the domain;

execute the at least one command associated with the user intent, wherein the at least one command is configured to direct the computing device to carry out at least one function selected from the set of functions; and

generate output of at least one result of the executed at least one command from at least one component of the computing device.

2. The computing device of claim 1, wherein the NLP engine selects at least one conditional random field (CRF) in response to the at least one domain.

3. The computing device of claim 2, wherein the instructions further configure the computing device to:

store at least one optimized feature set associated with the at least one domain, the at least one CRF configured to access the at least one feature set.

4. The method of claim 2, further comprising:

identifying at least one unfilled entity relating to the at least one command; and

generating at least one dialogue response comprising a prompt to obtain the at least one unfilled entity.

5. The computing device of claim 1, wherein the instructions further configure the computing device to:

assign one of the priority levels to one or more commands when the received at least one user query indicates more than one user intent for simultaneous multiple conversations, wherein each of the more than one user intents constitutes one conversation;

execute the one or more commands according to the assigned one of the priority levels;

assign a pending status to the one or more commands until completed; and

generate output indicating a pending status when the received at least one query is associated with the one or more commands with the pending status.

6. The computing device of claim 1, wherein the at least one command and at least one entity are provided to a services component for identifying at least one service capable of performing the at least one command, and for instructing the at least one service to perform the at least one command, the at least one service configured to return a service result to the services component in response to performing the at least one command.

7. The computing device of claim 1, wherein the instructions further configure the computing device to:

identify at least one dialogue response to the at least one query.

8. The computing device of claim 7, wherein the instructions further configure the computing device to:

format an output, the output comprising the at least one dialogue response.

9. The computing device of claim 8, wherein the at least one dialogue response is formatted in a conversational manner.

10. The computing device of claim 7, wherein the NLP engine is configured to create a list of candidate interpretations of the at least one user query, and wherein the at least one dialogue response is presented in the list for selection.

11. The computing device of claim 7, wherein the instructions further configure the computing device to:

identify at least one unfilled entity relating to the at least one command; and

generate at least one dialogue response comprising a prompt to obtain the at least one unfilled entity.

12. A method of processing a user query performed by a computing device, the method comprising:

receiving at least one user query from an input device, wherein the at least one user query includes a user intent and at least one command, the at least one user intent being associated with a set of functions to be carried out by the computing device;

deriving, using a natural language processing (NLP) engine, the a user intent from the user query;

identifying, using the NLP engine, at least one domain and the at least one command associated with the derived user intent, wherein the NLP engine performs entity extraction on the at least one user query to identify at least one entity associated with the at least one command and wherein the command is for performing by a service associated with the domain;

execute the at least one command associated with the user intent, wherein the at least one command is configured to direct the computing device to carry out a function selected from the set of functions; and

generate output of at least one result of the executed at least one command from at least one component of the computing device.

13. The method of claim 12, wherein performing the NLP engine comprises selecting at least one conditional random field (CRF) in response to the at least one domain identified.

14. The method of claim 13, further comprising storing at least one optimized feature set associated with the at least one identified domain, and wherein the at least one CRF is configured to access the at least one feature set.

15. The method of claim 13, further comprising:

identifying at least one dialogue response to the at least one query.

16. The method of claim 15, further comprising:

formatting an output, wherein the output comprises the at least one dialogue response.

17. The method of claim 15, further comprising:

defining the at least one dialogue response in a conversational manner.

18. The method of claim 12, further comprising:

identifying at least one service capable of performing the at least one command; and

instructing the at least one service to perform the at least one command, the at least one service configured to return a service result to the services component in response to performing the at least one command.

19. A computer storage device storing computer-executable instructions that when executed by a processor cause the processor to:

receive at least one user query from an input device, wherein the at least one user query includes a user intent and at least one command, the at least one user intent being associated with a set of functions to be carried out by the computing device;

derive, using a natural language processing (NLP) engine, the user intent from the user query;

identify, using the NLP engine, at least one domain and the at least one command associated with the derived user intent, wherein the NLP engine performs entity extraction on the at least one user query to identify at least one entity associated with the at least one command, wherein the at least one command is associated with the domain;

execute the at least one command associated with the user intent, wherein the at least one command is configured to direct the computing device to carry out at least one function selected from the set of functions; and

generate output of at least one result of the executed at least one command from at least one component of the computing device.

20. The A computer storage device of claim 19, the computer-executable instructions further causing the processor to:

identify at least one dialogue response to the at least one query.