METHODS AND SYSTEMS FOR RECOMMENDING DIALOGUE STICKER BASED ON SIMILAR SITUATION DETECTION

• US 20160210117A1
• Filed: 07/14/2015
• Published: 07/21/2016
• Est. Priority Date: 01/19/2015
• Status: Active Grant

• Current Assignee: NCSoft Corporation
• Original Assignee: NCSoft Corporation
• Inventors: Kim, Taek Jin, Lee, Jay June, Jang, Jungsun, Chung, Sehee, Lee, Kyeong Jong, Lee, Yeonsoo
• Granted Patent: US 9,792,909 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: H04L 51/52 for supporting social networking services
• CPC Class Codes: H04M 1/7243 with interactive means for internal management of messages

Abstract

Provided is a method of recommending a dialogue sticker through similar situation retrieval. The method includes: generating dialogue situation information on the last utterances of the first user terminal and the second user terminal; retrieving a candidate situation having the most similar context from a dialogue situation information database, the situation including a plurality of continuous utterances; acquiring sticker information from pairs of utterance-sticker derived from the retrieved candidate situation and providing the sticker information for the first user terminal.

Claims

1. A method of recommending a dialogue sticker through similar situation retrieval, comprising:

by a server connected to a first user terminal and a second user terminal through a network,

generating dialogue situation information on the last utterances of the first user terminal and the second user terminal;

retrieving a candidate situation having the most similar context from a dialogue situation information database, the situation comprising a plurality of continuous utterances;

acquiring sticker information from pairs of utterance-sticker derived from the retrieved candidate situation and providing the sticker information for the first user terminal.

2. The method of claim 1, wherein the dialogue situation information database comprises the pairs of utterance-sticker and dialogue situation information on the plurality of continuous utterances that are already analyzed.

3. The method of claim 1, wherein the generating of the dialogue situation information comprises generating dialogue situation information comprising dialogue act category information on a certain number of continuous utterances comprising the last utterance and exchanged between the first user terminal and the second user terminal.

4. The method of claim 1, between the retrieving of the candidate situation and the providing of the sticker information for the first user terminal, further comprising:

determining whether the first user terminal is in a speaker utterance situation or a response utterance situation; and

retrieving the speaker utterances from the candidate situation in case of the speaker utterance situation, and retrieving the response utterances in case of the response utterance situation.

5. The method of claim 1, wherein the retrieving of the candidate situation comprises retrieving a situation having continuous utterances matching with the flow of dialogue act of the inputted last utterance from the dialogue situation information database.

6. The method of claim 5, comprising determining the ranking of the candidate situation in accordance with a dialogue act flow prediction criteria that are already defined in regard to the retrieved similar situation,

wherein the providing of the sticker information for the first user terminal comprises preferentially providing sticker information acquired from a high-ranked candidate situation.

7. The method of claim 5, further comprising determining whether there is an utterance speaker switching when determining accordance of the flow of dialogue act in regard to the plurality of continuous utterances.

8. The method of claim 5, wherein the retrieving of the candidate situation comprises,

inferring a relationship between the first user and the second user from the dialogue situation information of the inputted last utterance and retrieving a situation matching with the relationship between the first user and the second user from the dialogue situation information database.

9. The method of claim of 1, after the generating of the dialogue situation information, further comprising:

determining whether or not the dialogue situation information that is analyzed meets event rules that are already defined; and

proceeding to retrieving of the candidate situation when the event rules are not met, and withdrawing sticker information stored in association with the corresponding rules for the first user terminal so as to provide the sticker information and finish the procedure when the event rules are met.

10. A computer program recorded with:

a dialogue situation information database storing dialogue situation information and pairs of utterance-sticker in regard to a plurality of continuous utterances that are already analyzed;

a sticker database storing information on a plurality of stickers; and

a plurality of commands for;

generating dialogue situation information on the last utterances of a first user terminal and a second user terminal;

retrieving a candidate situation—

in this case, the situation comprising a plurality of continuous utterances—

having the most similar context from a dialogue situation information database;

acquiring sticker information from pairs of utterance-sticker derived from the retrieved candidate situation and providing the sticker information for a first user terminal.

11. The computer program of claim 10, wherein the retrieving of the candidate situation comprises retrieving a situation having continuous utterances matching with the flow of dialogue act of the inputted last utterance from the dialogue situation information database.

12. The computer program of claim 11, determining the ranking of the candidate situation in accordance with dialogue act flow prediction criteria that are already defined in regard to the retrieved similar situation,

wherein the providing of the sticker information for the first user terminal comprises preferentially providing sticker information acquired from a high-ranked candidate situation.

13. The computer program of claim 11, further determining whether there is an utterance speaker switching when determining accordance of the flow of dialogue act in regard to the plurality of continuous utterances.

14. The computer program of claim 11, wherein the retrieving of the candidate situation comprises retrieving a situation matching with the relationship between the first user and the second user from the dialogue situation information database when inferring a relationship between the first user and the second user from the dialogue situation information of the inputted last utterance.