Dialogue management method, user terminal and computer-readable recording medium
View Patent ↗A dialogue management method includes: obtaining at least one of a speech recognition result or a natural language understanding result for a speech of a user; determining whether a fixed speech pattern is included in the speech of the user, based on the at least one of the speech recognition result or the natural language understanding result for the speech of the user; and when the fixed speech pattern is included in the speech of the user, generating recommended speech information for guiding a completion of the speech of the user, based on at least one of a current dialogue context or a type of the fixed speech pattern included in the speech of the user.
1. A dialogue management method, comprising:
obtaining at least one of a speech recognition result or a natural language understanding result for a speech of a user;
determining whether a fixed speech pattern is included in the speech of the user, based on the at least one of the speech recognition result or the natural language understanding result for the speech of the user; and
when the fixed speech pattern is included in the speech of the user, generating recommended speech information for guiding a completion of the speech of the user, based on at least one of a current dialogue context or a type of the fixed speech pattern included in the speech of the user, and
wherein the fixed speech pattern includes at least one of a first speech pattern including only a specific constituent element which is meaningless, or a second speech pattern including a meaningful constituent element and the specific constituent element positioned after the meaningful constituent element.
2. The dialogue management method of claim 1 , wherein the type of the fixed speech pattern varies depending on a position of a specific constituent element included in the speech of the user and a characteristic of another constituent element uttered before the specific constituent element.
3. The dialogue management method of claim 1 , wherein, when the first speech pattern is included in the speech of the user and a dialogue between a dialogue management system and the user is ongoing, the generating of the recommended speech information includes generating the recommended speech information based on an example speech corresponding to the ongoing dialogue.
4. The dialogue management method of claim 1 , wherein, when the second speech pattern is included in the speech of the user and a dialogue between a dialogue management system and the user is ongoing, the generating of the recommended speech information includes generating the recommended speech information by use of a constituent element omitted from the speech of the user among constituent elements of an example speech corresponding to the ongoing dialogue.
5. The dialogue management method of claim 1 , wherein, when the second speech pattern is included in the speech of the user and a dialogue between a dialogue management system and the user is not ongoing, the generating of the recommended speech information includes generating the recommended speech information based on an example speech similar to the speech of the user.
6. The dialogue management method of claim 5 , wherein, when the second speech pattern is included in the speech of the user and the dialogue between the dialogue management system and the user is not ongoing, the generating of the recommended speech information includes generating the recommended speech information by use of a constituent element omitted from the speech of the user among constituent elements of the example speech similar to the speech of the user.
7. The dialogue management method of claim 1 , wherein the obtaining of the at least one of the speech recognition result or the natural language understanding result for the speech of the user includes obtaining the at least one of the speech recognition result or the natural language understanding result for the speech of the user in real time.
8. The dialogue management method of claim 1 , wherein the specific constituent element includes a filler which is a meaningless constituent element input before, after or during speech.
9. The dialogue management method of claim 1 ,
wherein the meaningful constituent element included in the second speech pattern forms an incomplete speech, and
wherein the fixed speech pattern further includes a third speech pattern including a constituent element forming a complete speech and the specific constituent element positioned after the constituent element forming the complete speech.
10. The dialogue management method of claim 9 , further including:
when the third speech pattern is included in the speech of the user, generating a system response corresponding to the speech of the user.
11. A user terminal, comprising:
a microphone to which a speech of a user is input;
a communication module configured to transmit the speech of the user to a dialogue management system and receive recommended speech information for guiding a completion of the speech of the user from the dialogue management system; and
a display configured to display the received recommended speech information,
wherein the received recommended speech information is generated based on at least one of a current dialogue context or a type of a fixed speech pattern included in the speech of the user, and
wherein the fixed speech pattern includes at least one of a first speech pattern including only a specific constituent element which is meaningless, or a second speech pattern including a meaningful constituent element and the specific constituent element positioned after the meaningful constituent element.
12. The user terminal of claim 11 , wherein, when the first speech pattern is included in the speech of the user and a dialogue between the dialogue management system and the user is ongoing, the display is configured to display the received recommended speech information generated based on an example speech corresponding to the ongoing dialogue.
13. The user terminal of claim 11 , wherein, when the second speech pattern is included in the speech of the user and a dialogue between the dialogue management system and the user is ongoing, the display is configured to display the received recommended speech information generated by use of a constituent element omitted from the speech of the user among constituent elements of an example speech corresponding to the ongoing dialogue.
14. The user terminal of claim 11 , wherein, when the second speech pattern is included in the speech of the user and a dialogue between the dialogue management system and the user is not ongoing, the display is configured to display the received recommended speech information generated based on an example speech similar to the speech of the user.
15. The user terminal of claim 14 , wherein, when the second speech pattern is included in the speech of the user and the dialogue between the dialogue management system and the user is not ongoing, the display is configured to display the recommended speech information generated by use of a constituent element omitted from the speech of the user among constituent elements of the example speech similar to the speech of the user.
16. The user terminal of claim 11 , wherein the communication module is configured to transmit the input speech of the user to the dialogue management system in real time.
17. The user terminal of claim 11 , wherein the specific constituent element includes a filler which is a meaningless constituent element input before, after or during speech.
18. A computer-readable recording medium storing a program for implementing a dialogue management method, the dialogue management method comprising:
obtaining at least one of a speech recognition result or a natural language understanding result for a speech of a user;
determining whether a fixed speech pattern is included in the speech of the user, based on the at least one of the speech recognition result or the natural language understanding result for the speech of the user; and
when the fixed speech pattern is included in the speech of the user, generating recommended speech information for guiding a completion of the speech of the user, based on at least one of a current dialogue context or a type of the fixed speech pattern included in the speech of the user,
wherein the fixed speech pattern includes at least one of a first speech pattern including only a specific constituent element which is meaningless, or a second speech pattern including a meaningful constituent element and the specific constituent element positioned after the meaningful constituent element.
19. The recording medium of claim 18 , wherein, when the first speech pattern is included in the speech of the user and a dialogue between a dialogue management system and the user is ongoing, the generating of the recommended speech information includes generating the recommended speech information based on an example speech corresponding to the ongoing dialogue.
20. The recording medium of claim 18 , wherein, when the second speech pattern is included in the speech of the user and a dialogue between a dialogue management system and the user is ongoing, the generating of the recommended speech information includes generating the recommended speech information by use of a constituent element omitted from the speech of the user among constituent elements of an example speech corresponding to the ongoing dialogue.
21. The recording medium of claim 18 , wherein, when the second speech pattern is included in the speech of the user and a dialogue between a dialogue management system and the user is not ongoing, the generating of the recommended speech information includes generating the recommended speech information based on an example speech similar to the speech of the user.
22. The recording medium of claim 21 , wherein, when the second speech pattern is included in the speech of the user and the dialogue between the dialogue management system and the user is not ongoing, the generating of the recommended speech information includes generating the recommended speech information by use of a constituent element omitted from the speech of the user among constituent elements of the example speech similar to the speech of the user.
23. The recording medium of claim 18 , wherein the obtaining of the at least one of the speech recognition result or the natural language understanding result for the speech of the user includes obtaining the at least one of the speech recognition result or the natural language understanding result for the speech of the user in real time.
24. The recording medium of claim 21 , wherein the specific constituent element includes a filler which is a meaningless constituent element input before, after or during speech.
25. The recording medium of claim 21 ,
wherein the meaningful constituent element included in the second speech pattern forms an incomplete speech, and
wherein the fixed speech pattern further includes a third speech pattern including a constituent element forming a complete speech and the specific constituent element positioned after the constituent element forming the complete speech.
26. The recording medium of claim 25 , further including:
when the third speech pattern is included in the speech of the user, generating a system response corresponding to the speech of the user.