Method and system for searching for and providing information about natural language query having simple or complex sentence structure

Patent metadata

Field	Value
Publication number	US-2016012105-A1
Application number	US-201514796074-A
Country	US
Kind code	A1
Filing date	Jul 10, 2015
Priority date	Jul 10, 2014
Publication date	Jan 14, 2016
Grant date	—

Abstract

A method of searching for and providing information about a natural language query having a simple or complex sentence structure, includes: generating a mashup query language having a tree structure in a plurality of levels based on at least one query entity included in a natural language query language via a semantic analysis of the natural language query language; determining whether the plurality of levels are linked through a query entity forming each of the plurality of levels based on attribute information of the mashup query language; searching for data corresponding to the query entity forming each of the plurality of levels from a knowledge database based on a result of the determining, and deriving main information and at least one piece of entity information corresponding to the natural language query language from found data; and laying out a search result screen including the main information and the at least one piece of entity information.

First claim

What is claimed is:

1 . A method of searching for and providing information about a natural language query having a simple or complex sentence structure, the method comprising:
generating a mashup query language having a tree structure in a plurality of levels based on at least one query entity included in a natural language query language via a semantic analysis of the natural language query language; determining whether the plurality of levels are linked through a query entity forming the mashup query language of each of the plurality of levels based on attribute information of the mashup query language; searching for data corresponding to the query entity forming the mashup query language of each of the plurality of levels from a knowledge database based on a result of the determining, and deriving main information and at least one piece of entity information corresponding to the natural language query language from the searched data; and laying out a search result screen comprising the main information and the at least one piece of entity information.

2 . The method of claim 1 , wherein the generating of the mashup query language comprises:
determining whether the natural language query language has the simple sentence structure or the complex sentence structure via the semantic analysis of the natural language query language; when the natural language query language has the complex sentence structure, deriving at least two query phrases from the natural language query language; and generating the mashup query language having the tree structure based on the at least two query phrases.

3 . The method of claim 1 , wherein the determining of whether the plurality of levels are linked comprises:
analyzing the attribute information of the mashup query language and domain information of the mashup query language; selecting search environment information corresponding to the mashup query language according to a result of the analyzing; and normalizing the mashup query language by using the selected search environment information.

4 . The method of claim 3 , wherein the determining of whether the plurality of levels are linked further comprises converting the normalized mashup query language to an object query language by using the selected search environment information.

5 . The method of claim 1 , wherein, when it is determined that the plurality of levels are linked, the searching for the data corresponding to the query entity, comprises:
searching for and deriving data corresponding to an uppermost level entity of the mashup query language from the knowledge database; and searching for and deriving data corresponding to consecutive lower level entities from the knowledge database based on the data corresponding to the uppermost level entity, wherein the searching for the data corresponding to the query entity is repeated until data corresponding to a lowermost level entity of the mashup query language is searched for and derived.

6 . The method of claim 5 , wherein the searching for the data corresponding to the query entity further comprises deriving the data corresponding to the lowermost level entity as the main information and deriving the data corresponding to the remaining level entities as the at least one piece of entity information.

7 . The method of claim 6 , wherein the laying out of the search result screen comprises outputting the data corresponding to the uppermost level entity at a top region of the search result screen and outputting the data corresponding to the consecutive lower level entities below the data corresponding to the uppermost level entity, wherein the data corresponding to the lowermost level entity is output as the main information.

8 . The method of claim 1 , wherein the laying out of the search result screen comprises processing the main information and the at least one piece of entity information to compressed information comprising at least one of at least one piece of attribute information, image information, and video information.

9 . The method of claim 8 , further comprising:
when a user input of selecting one of the at least one piece of entity information is received on the search result screen, processing the selected piece of entity information as entity detail information; and changing a layout of the search result screen by dividing the search result screen into a compressed information region where the main information and the at least one piece of entity information are output, and a detail information region where the entity detail information is output.

10 . The method of claim 1 , wherein, when it is determined that the plurality of levels are not linked, the searching for the data corresponding to the query entity comprises:
deriving final result data by applying a logical operation according to a result of the semantic analysis on the data corresponding to the query entity of each of the plurality of levels.

11 . The method of claim 10 , wherein the searching for the data corresponding to the query entity further comprises deriving the final result data as the main information and the data corresponding to the query entity as the at least one piece of entity information.

12 . The method of claim 11 , wherein the laying out of the search result screen comprises laying out the search result screen by dividing the search result screen into a main information region where the main information and main information-related information are output, and an entity region where the at least one piece of entity information is output.

13 . The method of claim 1 , wherein the laying out of the search result screen comprises, when the main information is single attribute information, processing and providing the main information to a form of a natural language based response.

14 . A system for searching for and providing information about a natural language query having a simple or complex sentence structure, the system comprising:
a natural language process engine configured to generate a mashup query language having a tree structure in a plurality of levels based on at least one query entity included in a natural language query language input through a user device via a semantic analysis of the natural language query language; a search engine configured to determine whether the plurality of levels are linked based on attribute information of the mashup query language and search for data corresponding to a query entity forming each of the plurality of levels, wherein the search engine processes, from the data corresponding to the query entity, main information and at least one piece of entity information, which correspond to the natural language query language, and displays the main information and the at least one piece of entity information on a same search result screen.

15 . The system of claim 14 , wherein the natural language process engine comprises:
a semantic analyzing module configured to determine whether the natural language query language has the simple sentence structure or the complex sentence structure via the semantic analysis of the natural language query language, and when the natural language query language has the complex sentence structure, derive at least two query phrases from the natural language query language; and a mashup query language converting module configured to generate the mashup query language based on the at least two query phrases.

16 . The system of claim 14 , wherein the search engine comprises:
a mashup query language normalizing module configured to analyze attribu

Assignees

• Naver Corp

Inventors

• Chang Jae Hyeok
• Kim Ki Young
• Kim Myeong Seok
• Choi Ji Hye
• Lee Won Jin
• Lee Hyun Ah
• Lee Yong Hun

Classifications

• G06F16/248

  Presentation of query results · CPC title

• G06F16/245

  Query processing · CPC title

• G06F16/243 Primary

  Natural language query formulation · CPC title

• G06F16/24534

  Query rewriting; Transformation · CPC title

• G06F16/43 Primary

  Querying · CPC title