System for processing user voice utterance and method for operating same

1. An apparatus comprising:

a display;

at least one memory; and

at least one processor operatively connected to the at least one memory,

wherein the at least one memory stores instructions that, when executed, cause the at least one processor to:

control the display to display a user interface to select at least one previously selected domain,

based on the user interface, receive a user input to select the at least one previously selected domain,

receive an input corresponding to a user utterance,

recognize content from the user utterance,

identify whether the content includes an explicit service domain name corresponding to at least one domain to determine an intent of the user utterance,

based on identifying that the explicit service name is included in the content,

perform a function corresponding to the at least one domain corresponding to the explicit service domain name, and

based on identifying that no explicit service name is included in the content:

identify whether the user utterance corresponds to the at least one domain previously selected by the user,

when the user utterance corresponds to the at least one domain previously selected by the user, perform a function corresponding to the at least one domain previously selected by the user, and

when the user utterance does not correspond to the at least one domain previously selected by the user, generate response information to provide to user based on at least one domain selected from among a plurality of normal service domains by predicting based on the user utterance,

wherein the explicit service domain name refers to an entity providing goods and/or services to the user.

2. The apparatus of claim 1 , wherein the instructions, when executed, further cause the at least one processor to determine whether the content includes the explicit service domain name by using a first natural language understanding model related to multiple domains.

3. The apparatus of claim 2 , wherein the first natural language understanding model comprises at least one of a domain determination model or an intent determination model.

4. The apparatus of claim 3 , wherein the instructions, when executed, further cause the at least one processor to:

analyze the content based on the domain determination model to determine a confidence score, and

determine whether the content includes the explicit service domain name based on the confidence score.

5. The apparatus of claim 4 , wherein the instructions, when executed, further cause the at least one processor to:

determine a second natural language understanding model corresponding to the at least one domain included in the content, and

determine at least one of a user intent or a parameter by using the second natural language understanding model.

6. The apparatus of claim 5 , wherein the instructions, when executed, further cause the at least one processor to:

based on the user input, train the first natural language understanding model.

7. The apparatus of claim 6 , wherein the instructions, when executed, further cause the at least one processor to:

acquire rule information corresponding to the at least one domain previously selected by the user, and

train the second natural language understanding model by using the rule information.

8. The apparatus of claim 1 , wherein the at least one domain previously selected by the user is selected from a list of multiple domains classified according to a selected standard.

9. The apparatus of claim 1 , further comprising a user interface comprising:

a guide to select the at least one domain previously selected by the user, and

a service that can be provided to the user by using the at least one domain previously selected by the user.

10. The apparatus of claim 1 , wherein the apparatus comprises a mobile terminal, a stationary terminal, or a server.

11. The apparatus of claim 10 , wherein the instructions further cause the at least one processor to:

in the case that the content includes a business entity, determine whether a first confidence score is greater than a first threshold associated with a first domain of the at least one domain included in the content and determine whether a second confidence score is greater than a second threshold associated with a second domain, and

in the case that the first confidence score is greater than the first threshold and the second confidence score is not greater than the second threshold, process the content using the first domain.

12. The apparatus of claim 1 , wherein the at least one domain is related to a type of a service provided to the user to perform a user intent included in the user utterance or a subject that provides the service.

13. The apparatus of claim 1 , wherein the at least one domain corresponds to a service provider or a service capsule corresponding to a service.

14. A method for operating an apparatus, the method comprising:

controlling a display to display a user interface to select at least one previously selected domain;

based on the user interface, receiving a user input to select the at least one previously selected domain;

receiving an input corresponding to a user utterance;

recognizing content from the user utterance;

identifying whether the content includes an explicit service domain name corresponding to at least one domain to determine an intent of the user utterance;

based on identifying that the explicit service domain name is included in the content, performing

a function corresponding to the at least one domain corresponding to the explicit service domain name; and

based on identifying that no explicit service name is included in the content:

identifying whether the user utterance corresponds to the at least one domain previously selected by the user,

when the user utterance corresponds to the at least one domain previously selected by the user, performing a function corresponding to at least one domain previously selected by the user, and

when the user utterance does not correspond to the at least one domain previously selected by the user, generating response information to provide to user based on at least one domain selected from among a plurality of normal service domains by predicting based on the user utterance,

wherein the explicit service domain name refers to an entity providing goods and/or services to the user.

15. The method of claim 14 , wherein the identifying of whether the content includes the explicit service domain name comprises determining whether the content includes the explicit service domain name corresponding to the at least one domain or a service capsule corresponding to the service by using a first natural language understanding model related to multiple domains.

16. The method of claim 15 , wherein the first natural language understanding model comprises at least one of a domain determination model or an intent determination model.

17. The method of claim 16 , wherein the determining of whether the content corresponds to the at least one domain comprises:

analyzing the content based on the domain determination model to determine a confidence score; and

identifying whether the content includes the at least one domain based on the confidence score.

18. The method of claim 17 , wherein the performing of the function corresponding to the at least one domain previously selected by the user comprises:

determining a second natural language understanding model corresponding to the at least one domain previously selected by the user; and

determining at least one of a user intent or a parameter by using the second natural language understanding model.

19. The method of claim 16 , further comprising:

based on the user input, training the first natural language understanding model.

20. The method of claim 19 , wherein the training of the first natural language understanding model comprises:

acquiring rule information corresponding to the at least one domain previously selected by the user; and

training the first natural language understanding model by using the rule information.

21. The method of claim 20 , wherein the training of the first natural language understanding model comprises:

generating training data based on use history information; and

applying the training data to the first natural language understanding model based on the rule information.

22. The method of claim 14 , wherein the at least one domain previously selected by the user from a list of multiple domains classified according to a selected standard.

23. The method of claim 14 ,

wherein the at least one domain previously selected by the user is related to at least one object included in a user interface, and

wherein the user interface comprises:

a guide to select the at least one domain previously selected by the user, and

a service that can be provided to the user by using the at least one domain previously selected by the user.

24. The method of claim 14 , further comprising:

in the case that the content includes a business entity, determining whether a first confidence score is greater than a first threshold associated with a first domain of the at least one domain included in the content and determining whether a second confidence score is greater than a second threshold associated with a second domain; and

in the case that the first confidence score is greater than the first threshold and the second confidence score is not greater than the second threshold, processing the content using the first domain.

25. The method of claim 14 , wherein the at least one domain is related to a type of a service provided to the user to perform a user intent included in the user utterance or a subject that provides the service.

26. The method of claim 14 , further comprising:

in the case that the content comprises a business entity:

determining a user intent and a parameter associated with the user intent based on a first natural language understanding model associated with the business entity, and

performing an action based on the user intent and the parameter.

27. The method of claim 26 , further comprising receiving the first natural language understanding model from the business entity or a third party that is designated to provide the model on behalf of the business entity.