KR101602342B1 - Method and system for providing information conforming to the intention of natural language query - Google Patents

Abstract

The present invention relates to a method of extracting and providing information in accordance with a natural language query intention, comprising: converting a semantic-tagged natural language query into a mesh-up search language; Selecting search environment information corresponding to the converted mesh-up search language based on attribute information of the mesh-up search language, and converting the converted mesh-up search language into a target search language using the selected search environment information ; Retrieving data corresponding to the converted target search language from a knowledge database according to the search environment information; Processing the retrieved data according to the retrieval environment information, and laying out a retrieval result screen including the processed data, wherein the retrieval environment information includes at least one of a retrieval mode defined by a query type, A search condition, and a search result providing method set for each search property defined by a method of deriving a result value for each search domain and attribute defined as a search field.

Description

Field of the Invention [0001] The present invention relates to a method and system for extracting and providing information in accordance with the intention of a meaningful tagged natural language query,

More particularly, the present invention relates to a method and system for laying out an information extraction and a search result screen by applying various information extraction methods matching user intention to a semantic-tagged natural language query, .

Ontology defines a semantic relationship between real-world objects and provides semantic specification of information for the computer to access information derived from complex relationships between objects. It is used for artificial intelligence, information retrieval, ubiquitous, e-commerce, etc. It can be applied in various prefectures.

In particular, the ontology expresses a variety of information on the web in a semantic relationship between objects, thereby obtaining a logical query result that can not be found in an existing keyword-based search in the information search field. For example, when using an ontology, you can get the results of a query such as "Where is the population of more than 70 million, and where is the capital of the neighboring country?"

Such ontology information retrieval is more useful in three aspects than keyword based retrieval. First, the keyword-based search results provide only the documents containing the input keywords as search results. However, the ontology information search can directly provide the specific information that the user wants to search and attribute information thereof. Second, keyword based search can not guarantee the search result reflecting the relation between input keywords. However, the ontology information search can find information from the relation with other objects that can characterize the search target. Third, keyword - based search can search only existing information, but searching ontology information can deduce new information that did not exist in the past by searching for a new relation from the relationship between objects.

However, this ontology information retrieval is a great help in user's convenience by directly searching for a query object desired by the user, but it is difficult to access the ontology query language because it is not familiar to the general user. As a result, the ontology becomes ineffective because the user has to learn the structure of the ontology and the language of the ontology to access the information constructed by the ontology.

On the other hand, natural language is enough to express the logical meaning expressed in ontology query language, and is familiar to the general user. Therefore, if the ontology information can be accessed by the natural language query, it is possible to improve the efficiency of the ontology information search by making the access of the general user easy.

In order to access ontology information by natural language query, it is required to convert natural language query into ontology query language. To do this, first, structure and meaning of natural language query are analyzed by using natural language processing technology, We need to find the object and relation of the ontology corresponding to the natural language expression and convert the natural language query into the ontology query language according to the grammar structure of the ontology query language.

However, this method has a problem that an ontology query language which can not be accessed due to errors such as structural analysis and semantic analysis on a natural language query occurs. More specifically, there is no guarantee that general natural language processing techniques have all the language resources needed to process natural language expressions that can access the ontology to be searched, and it can lead to false analysis results due to minor grammatical errors. An ontology query term that can not be generated may occur.

As a result, according to the existing retrieval method, the user has to learn the types of queries that the system can process while experiencing many trial and error about an error query. As a result, the user who uses the ontology- It does not know what query is useful because it does not know the type of query that the processing technology can accept and the specific content of the information constructed in the ontology.

Also, from the service provider 's point of view, it means that it can not provide all of the constructed ontology information.

However, most users want to intuitively use the information on the ontology prepared by the service provider rather than knowing the information about the query and the method of inputting the query. As a result, , The utilization and satisfaction of the system is reduced.

Therefore, research on Natural Language Understanding (NLU) is continuing to improve the utility and accuracy of information retrieval. Natural language technology is intended to grasp the user's intention of the user on the 'user question'. It is the input of the question which is uttered by voice or inputted as text, and the extracted list by understanding the intention of the question in the natural language technique, .

Currently, natural language based retrieval system is studying the way of providing retrieval result by grasping user 's intention of query contained in query language, unlike keyword based retrieval.

DISCLOSURE OF THE INVENTION An object of the present invention is to solve the above-mentioned problems, and it is an object of the present invention to provide a method of extracting various information according to mode, domain, property, And to propose a method and system for extracting information according to the user's intention according to the application.

It is another object of the present invention to provide a retrieval information providing method and system for laying out and providing a retrieval result screen so that a retrieval result for a user query is matched with a query intention.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are not restrictive of the invention, unless further departing from the spirit and scope of the invention as defined by the appended claims. It will be possible.

According to an aspect of the present invention, there is provided a method of searching natural language queries, the method comprising: (a) converting a semantic-tagged natural language query into a mesh-up search language; (b) selecting search environment information corresponding to the converted mesh-up search language based on the attribute information of the mesh-up search language, and selecting the converted mesh-up search language as a target search language Converting; (c) retrieving data corresponding to the converted target search language from a knowledge database according to the search environment information; (d) processing the retrieved data according to the retrieval environment information, and laying out a retrieval result screen including the processed data, wherein the retrieval environment information includes at least one of a retrieval mode defined by a query type, A search condition, and a search result providing method set according to a search property defined by a method of deriving a result value for each search domain and an attribute defined as an included search field.

The search mode according to an embodiment of the present invention may define at least one of a maximum number of search data, a number of attributes for search, and a number of attribute usage for each attribute according to a query type, Wherein the search property defines at least one of an exposure attribute, an image providing method, an alignment method, and a link generation method, the search property including at least one of a search result type, a result domain, a structure of a target search language corresponding to a mesh- At least one of the default exposure properties can be defined.

The step (b) according to an embodiment of the present invention may include analyzing attribute and domain information of the mesh-up search language; Selecting search environment information corresponding to the mesh-up search language according to the analysis result; Normalizing the mesh-up search language using the selected search environment information; And converting the normalized mesh-up search language into the target search language using the selected search environment information.

According to an embodiment of the present invention, the step (c) includes: deriving one or more physical data corresponding to the converted target search language from the knowledge database according to the search environment information; And deriving result data from the derived one or more physical data using the logic operation rule included in the search environment information.

According to an embodiment of the present invention, when the search mode is the correct type, the search environment information searches basic data corresponding to the user query and one or more additional data related to the main data, One or more additional data may be laid out on the same search result screen.

According to another embodiment of the present invention, when the search mode is a comparative type, the search environment information searches data of the same attribute for each of the comparison objects included in the user query term, Can be laid out on the same search result screen.

According to an embodiment of the present invention, when the search mode is graphical, the search environment information may include basic data for a search object matching the user query and one or more additional data related to the search object And define that the basic data and the one or more additional data are laid out in a graph form along with the relationship on the search result screen.

According to another aspect of the present invention, there is provided a natural language query-based retrieval system for normalizing a morpheme constituting a natural language query input through a user device, A natural language processing engine for converting a natural language query into a mesh-up search language; Up search environment corresponding to the converted mesh-up search language based on the attribute information of the mesh-up search language, and converts the converted mesh-up search language into a target search language, which is a database search language, based on the selected search environment information A search engine for performing a data search by converting the search data and processing the search data according to the search environment information; And a knowledge database that stores a plurality of search environment information, language conversion information, and knowledge information.

The search engine according to an embodiment of the present invention analyzes the attribute information of the mesh-up search language and selects search environment information corresponding to the mesh-up search language, and searches the mesh- A mesh-up search language normalization module that normalizes the language; A target search language conversion module for converting the normalized mesh-up search language into the target search language according to the search environment information; A data derivation module for retrieving and deriving physical data for the target search language from the knowledge database according to the search environment information to generate result data; And a data processing module for processing the result data generated by the data derivation module according to the search environment information and generating a search result screen including the processed data.

The data derivation module according to an embodiment of the present invention may derive one or more physical data corresponding to the converted target search language from the knowledge database according to the search environment information, The result data can be derived using the logical operation rules included in the environment information.

According to an embodiment of the present invention, when the search mode is the correct answer type, the data derivation module extracts, from the knowledge database, basic data conforming to the user query term and one or more additions And the data processing module may lay out a search result screen so that the basic data and the one or more additional data are included in the same screen according to the search environment information.

According to another embodiment of the present invention, when the search mode is a comparative type, the data derivation module derives data of the same attribute from each of the comparison objects included in the user query term from the knowledge database according to the search environment information , The data processing module may layout the search result screen so that the attribute data of each of the searched comparison objects is included in the same screen according to the search environment information.

According to another embodiment of the present invention, when the search mode is a graph type, the data derivation module extracts basic data for a search object matching the user query word from the knowledge database according to the search environment information, And the data processing module extracts the basic data and the one or more additional data according to the search environment information so that the basic data and the one or more additional data are displayed in a graph form on the screen of one search result Search result screen layout can be done.

The natural language processing engine according to an embodiment of the present invention may include: a normalization module for analyzing and normalizing a plurality of morphemes constituting the natural language query; And a semantic analysis module for extracting semantic expressions based on an Object Property Action Condition (OPAC) frame for the normalized morpheme and determining the normalized morpheme dependency relationship according to the extracted semantic expression.

At this time, the semantic analysis module constructs a dependency tree composed of at least two or more of at least one query object, at least one query attribute, and one or more query contents according to the normalized morpheme dependency relationship, You can create a mesh-up search language.

It is to be understood that both the foregoing general description and the following detailed description of the present invention are exemplary and explanatory and are intended to provide further explanation of the present invention by those skilled in the art. And can be understood and understood.

According to the embodiment of the present invention, by analyzing the semantic-tagged natural language query and applying various information extraction schemes according to the mode, domain, and property of the mesh-up search language, Information can be extracted to improve the accuracy and efficiency of retrieval.

In addition, according to the embodiment of the present invention, the search result screen for the user query word is laid out in accordance with the query intention to provide the search result screen, thereby increasing the user's interest and utilization.

BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of the specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention.
1 is a diagram showing an example of a semantic-tagged natural language query-based retrieval system according to an embodiment of the present invention.
2 is a flowchart illustrating an example of a process of extracting information and providing information in a selective manner according to attributes of a mesh-up search language in a search engine according to an embodiment of the present invention.
3 is a diagram showing an example of a mesh-up search language converted from a natural language query according to an embodiment of the present invention.
FIG. 4 is a diagram illustrating an example of normalization by applying search environment information (NQL config) according to an embodiment of the present invention.
FIG. 5 is a diagram showing an example of layout of data detected according to search environment information (NQL config) reflecting a user query intention according to an embodiment of the present invention.
FIG. 6 is a diagram illustrating another example of data laid out according to search environment information (NQL config) reflecting user query intention according to an embodiment of the present invention.
FIG. 7 is a diagram illustrating another example in which data detected according to search environment information (NQL config) reflecting a user query intention according to an embodiment of the present invention is laid out.
FIG. 8 is a diagram showing another example in which data detected according to search environment information (NQL config) reflecting a user query intention according to an embodiment of the present invention is laid out.

BRIEF DESCRIPTION OF THE DRAWINGS The present invention is capable of various modifications and various embodiments, and particular embodiments are illustrated in the drawings and described in the following description of the invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the accompanying drawings. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following detailed description, together with the accompanying drawings, is intended to illustrate exemplary embodiments of the invention and is not intended to represent the only embodiments in which the invention may be practiced. The following detailed description includes specific details in order to provide a thorough understanding of the present invention. However, those skilled in the art will appreciate that the present invention may be practiced without these specific details.

More particularly, the present invention relates to a method and system for laying out an information extraction and a search result screen by applying various information extraction methods matching user intention to a semantic-tagged natural language query, .

1 is a diagram showing an example of a semantic-tagged natural language query-based retrieval system according to an embodiment of the present invention.

Referring to FIG. 1, a search system 100 according to an embodiment of the present invention includes a natural language processing engine (NLU) based on a Natural Language Understanding (NLU) for analyzing a user query input through a user device 110 120, a search engine 130, and a knowledge database 140.

The natural language processing engine 120 analyzes a complex query term based on the semantically tagged natural language inputted through the user device 110 and converts it into a Naver Contents Repository-mashup Query Language (NQL) through semantic analysis, The search engine 130 derives the query-requested information from the pre-built knowledge database based on the NQL generated by the natural language processing engine 120 and provides it to the user device 110.

Hereinafter, each configuration of the search system according to the embodiment of the present invention will be briefly described.

The natural language processing engine 120 includes a semantic analysis module 121 for performing semantic analysis of a user query and an NQL conversion module 122 for converting a natural language-based user query into an inferable NQL according to a semantic analysis result.

The semantic analysis module 121 analyzes the user query word, classifies it into morpheme units, and analyzes the logical / physical double structure of the user query word by grasping the semantic relation between morphemes.

For example, if the user query term is "current age of a specific person A ?, ", the corresponding query term is a query phrase for requesting physical data" And the result of the query that requests the physical data of "Year?" Can be analyzed. In other words, through analyzing the semantic relationship between the morphemes of the query words, the "current age" can be understood as a logical / physical dual structure including the physical data such as "date of birth" and "current year" internally.

The NQL conversion module 122 is a module for converting a user query term based on a salmon into a search language NQL and can infer one or more query phrases included in the user query based on the analysis result performed by the semantic analysis module 121 NQL of the tree structure is created.

In this specification, NQL is a search language for specifying a service unit attribute by analyzing a query word, and can be defined as a language designed to query a knowledge database by meaning tagging a user query expressed in a natural language. NQL consists of a dual structure that externally represents logical attributes and internally generates search results using the physical attributes of the knowledge database. In other words, NQL is a language for building a unique logical structure that does not conform to the physical structure of the knowledge database, and defines the relationship between the logical structure and the physical structure through the setting so that appropriate data can be provided for each type of query and service type. For example, if the user query includes the phrase "how old are you?", The converted NQL can be specified as "age".

The NQL of the tree structure generated by the NQL conversion module 122 is as illustrated in FIG. This will be described later with reference to Fig.

Next, the search engine 130 performs information search from the knowledge database 140 based on the natural language based NQL generated by the natural language processing engine 120, and is based on the mesh-up technology.

Next, the search engine 130 reconfigures NQL for use in accordance with the query intention in consideration of query language frame analysis result search environment information (NQL config) converted into NQL based on the natural language processing engine 120 . In this specification, NQL config refers to the NQL attribute analysis, taking into account NQL mode, domain, and property, and considering the query intention divided for each service, Means search environment information that defines a data processing method.

Specifically, the search engine 130 includes an NQL normalization module 131 for applying NQL config corresponding to the NQL attribute to the NQL transmitted from the natural language processing system based on the knowledge database 140, an NQL normalizing module 131 for re- a data derivation module 133 for retrieving and deriving data corresponding to the OQL based on the knowledge database 140, and the retrieved retrieval data according to the user's intention And a data processing module 134 for processing the search result screen and laying out a search result screen.

The knowledge database 140 includes an NQL config DB 141 for storing various NQL config information according to the NQL attribute, an OQL DB 142 for storing attribute information and rule information required for conversion from NQL to OQL, And a data processing rule DB 144. [ At this time, the NQL config DB 141 stores various types of configs that define a data retrieval method conforming to the query intention in consideration of the NQL normalization according to the NQL attribute in consideration of the search mode, the search domain, and the search property.

The NQL normalization module 131 analyzes the attributes and domains of the NQL transmitted from the natural language processing engine to derive an NQL config corresponding to the query intention from the NQL config DB 141 and applies the derived NQL config to perform NQL normalization .

Table 1 illustrates NQL config according to an embodiment of the present invention.

NQL Config NQL config_mode 1) mode means user query type.
2) mode example: correct answer, comparison type, relational type, graph type, etc.
3) NQL Defines the maximum number of search values, the number of attributes used for search, and the number of attributes used for results. NQL config_domain 1) domain represents each category in multiple semantic hierarchies, representing the semantic category to which the query object or attribute belongs.
2) domain example: people, movies, broadcasting, etc.
3) Define basic exposure attributes, image provisioning method, sorting method, link generation method and so on. NQL config_property 1) property refers to the unit that describes the relationship or object between query objects.
2) property Example: age, family, movies, etc.
3) Define the kind of search value, domain of search value, OQL schema corresponding to NQL, processing method of search value, and basic exposure attribute.

Referring to Table 1, NQL config_mode means a user query type, which can be defined as the maximum number of retrieval values, the number of retrieval attributes, and the number of retrieval attributes for each attribute according to the query type. For example, the query type can be divided into correct answer type, comparison type, relational type, graph type, and the like. For each type, NQL config_mode for the correct answer query is defined as retrieving and providing the data accuracy corresponding to the query as a priority, and NQL config_mode for the comparison query is defined as The NQL config_mode for the relational query is defined to additionally retrieve data for another query object related to the two query objects, and the graph type The NQL config_mode for the query term can be defined as providing data corresponding to the query in the form of a graph.

NQL config_domain represents the semantic category to which the query or attribute belongs. The query language can consist of multiple semantic layers, each of which becomes a domain. For example, the query term "A movie starring A actress?" Can be composed of a character domain in the first hierarchy and a movie domain in the second hierarchy. NQL config_domain defines the basic exposure attributes, image provisioning method, sorting method, link generation method, etc. according to the attributes of each domain and restricts the conditions that one layer can have. This is to limit the number of different cases that can be present in the layer to provide more accurate and higher quality data.

NQL config_property is a unit that describes the relationship or object of a query object. It defines how to retrieve search data by NQL attribute. The manner of fetching the color data may include at least one of the kind of the result value, the domain of the result value, the OQL structure corresponding to the NQL, the processing method of the result value, and the basic exposure attribute.

Specifically, NQL config_property can be classified into a primitive type, an object type, an integral type molding, and a virtual molding type according to the NQL attribute. Among them, the virtual speed molding is a method that combines, computes, and processes physical data stored in a knowledge database to derive a result value, and defines a separate logic for the result. For example, in the case where the query language queries the "debian age" of a specific person, the attribute "debian age" Attribute.

NQL config_mode, NQL config_domain, and NQL config_property can be overridden with each other. For example, if the query has a default exposure attribute called "person domain" for "A movie starring actress?", But the resulting personality data provided through the "starring actor" attribute can set the movie-related attribute as the default exposure attribute. have. For example, when searching for "A actress" and searching for "B actress of the movie", the result attributes provided by default are the age of actress A, Occupation, education, etc. The latter result can be composed of the actress's age, occupation, and movies. That is, the result value provided according to the NQL attribute may differ in some part.

Referring again to FIG. 1, the NQL normalization module 131 converts the generated NQL based on the user query into a normalized NQL (NQL_normalization) by applying an NQL config corresponding to the query intention.

Preferably, the NQL normalization module 131 may include the clue information for deriving additional data for approximating the user query intention according to the NQL config while configuring NQL_normalization.

The OQL transformation module 132 converts the NQL normalization module 131 into OQL based on the NQL config based on the NQL normalization stored in the OQL DB 142 for the changed NQL normalization. At this time, the OQL transformation module 132 converts the desired data into the OQL that can derive the desired data as the physical property value of the knowledge information DB 143 based on the logical domain and the attribute of the NQL.

The data derivation module 133 searches the knowledge information DB 143 based on the converted OQL in the OQL transformation module 132 to derive the physical data.

The data derivation module 133 may convert the physical data corresponding to the OQL derived from the knowledge information DB 143 into the NQL result data according to the NQL config. In the resulting data conversion process, the physical structure results can be matched to the logical structure results.

At this time, the data derivation module 133 may derive the result data conforming to the logical structure by using one or more physical data derived from the knowledge information DB 143 when the NQL attribute is a virtual attribute.

The data processing module 134 processes the data in a format that conforms to the usage destination in which the data is used. Specifically, the query result data derived from the data derivation module 133 can be processed in accordance with the NQL config used for data retrieval in the current search engine, have. At this time, the data processing can process the NQL result data by fetching the data processing rules conforming to the NQL config from the data processing rule DB 144. For example, the layout of the data processing and the search result screen can be configured based on the basic exposure attribute, image providing method, data sorting method, link generating method, and the like defined according to NQL config_domain.

2 is a flowchart illustrating an example of a process of extracting information and providing information in a selective manner according to attributes of a mesh-up search language in a search engine according to an embodiment of the present invention.

Referring to FIG. 2, a search engine according to an embodiment of the present invention analyzes attributes and domains of NQL with respect to NQL transmitted from a natural language processing engine (S201).

According to the analysis result, the search engine derives NQL config corresponding to the user query intention from the NQL config DB, and normalizes the NQL received from the natural language processing engine to NQL_normalization based on the NQL config (S202).

NQL config is a data detection method defined according to the characteristics of a query mode, a query domain, and a query property based on an attribute of a query word, as described above with reference to Table 1. The NQL config, which is applied according to a user query intention, It can be different.

In this case, NQL_normalization can be configured by further including an additional data search item for searching for data optimized for the user query intention according to the NQL config attribute applied to the corresponding NQL. For example, if the information about a specific person (e.g., a feature film) is a query attribute, additional data search items can be set so that information about other characters appearing in the movie can be also provided.

When the NQL normalization is performed, the search engine converts the NQL_normalization to OQL, which is a search language for querying the knowledge database, according to the NQL config applied in the previous step S202 (S203). OQL uses the physical structure that the knowledge database uses.

In addition, in the process of data derivation, the maximum number of retrieval values, the number of retrieval attributes, and the attributes for results are used according to the query type according to the NQL config_mode in retrieving the knowledge database using the converted OQL And the like (S204).

For example, for a correct query (eg, the key of a particular person A), NQL config can be configured to first derive the physical data corresponding to the query. At this time, data on another person having the same data as the search result data may be derived according to the additional data search items included in the NQL_normalization, and the additional information may be also provided.

As another example, it is possible to derive data for attributes that are useful for comparison among the respective data corresponding to the comparison object, with respect to a comparative query term (e.g., A movie vs. B movie?).

Another example is to derive additional data for other query objects that relate to the query object for a relational query (eg, what is the identity of a particular person?), Calculate the relational distance between each object, Effective data can be derived.

Also, based on the search condition defined for each domain property according to NQL config_domain, the knowledge database can be searched based on data type, data domain, basic exposure attribute, etc. according to search according to NQL config_property.

Next, the OQL data derived in the previous step S204 is converted into NQL result data in a predetermined data processing manner according to the NQL config (S205). At this time, the result of the physical structure can be matched to the logical structure. For example, when the query term asks the "current age" of the person to be searched, the NQL result data is derived as a logical structure for deriving the physical data "birth date" and "current year" of the person and calculating "current age" can do.

As the NQL result data is derived, the search engine applies a data processing rule conforming to the NQL config to form a search result screen in a search result providing mode reflecting the user's intention and provides the layout to the user (S206). The layout of the search result screen for search result data can be processed differently for each question type based on NQL config_mode.

3 is a diagram showing an example of an NQL converted from a natural language query according to an embodiment of the present invention.

Referring to FIG. 3, the user query term is "What is the movie in which the member of the Girls 'Generation appeared? &Quot;, and the natural language processing engine searches the query term by" Girls' , And can be expressed as an NQL having a hierarchical structure as shown in FIG. 3 by constructing a dependency tree of "Girls Generation ← member ← movie" according to dependency between morphemes. The illustrated hierarchical structure can be composed of "person / member" 302 and "person / mpvie_performed" 303 as "object / name" The NQL attribute shown in FIG. 3 can be regarded as a hierarchical structure in the form of OPP (Object / Property / Property).

According to an embodiment of the present invention, the mesh-up technique includes an OP (Object / Property), a COP (Condition / Object / Property), and an OPP (Object / Property / Property) for each attribute of a morpheme constituting a query word, Such as OOP (Object / Object / Property), OO (Object / Object), CCO (Condition / Condition / Object) and CCOP (Condition / Condition / Object / Property).

FIG. 4 is a diagram illustrating an example of normalization by applying an NQL config conforming to a user's query intention according to an embodiment of the present invention. Referring to FIG.

Referring to FIG. 4, based on the attribute of the NQL converted from the user query term "What is a movie starring Girls Generation member?" Illustrated in FIG. 3, NQL is set so that necessary data can be detected in terms of the character domain and the movie domain It can be converted into a normalized NQL_normalization (401).

5 is a view showing an example of layout of data detected according to an NQL config reflecting a user query intention according to an embodiment of the present invention.

Referring to FIG. 5, when the user query term is a " specific singer ", the search engine according to the embodiment of the present invention selects the query type from the NQL attribute corresponding to the user query term " Lee Hyori & , And "person domain". Accordingly, in setting NQL config corresponding to the query, NQL config_mode (correct answer) and NQL config_domain (person domain) can be set.

That is, by using the maximum number of retrieval data defined in NQL config_mode (correct answer), the number of retrieval attributes used, and the number of retrieved attributes, the default exposure attribute, the image providing method, and the sorting method defined in NQL config_domain NQL config_property uses the defined conditions such as data type, data domain, OQL structure corresponding to the corresponding NQL, result data processing method, and basic exposure attribute to be derived for each attribute according to such setting. Can be used.

Accordingly, the search result screen can be laid out as the result data as shown in FIG. The profile photograph 501, the basic information 502, and the activity information 503 of the person are defined as the basic exposure attribute and the detailed information included in each attribute is listed with respect to the person query word . At this time, the method of outputting the detailed information can be variously configured as defined in NQL config_property. For example, when the job of the search target character has the attribute of "singer ", the type of information provided in the activity information 520 may include detailed information such as album, broadcast, performance, book, movie, It is possible to lay out the data processing and search result screen by applying NQL config defined that the album information is outputted first and each album is displayed in the form of image.

6 is a diagram showing another example of data laid out according to an NQL config reflecting a user query intention according to an embodiment of the present invention.

Referring to FIG. 6, when the user query term is "specific movie (Star Wars) ", the search engine according to the embodiment of the present invention extracts, from the NQL attribute corresponding to the user query term" Star Wars " "And corresponds to" movie domain ". Accordingly, in setting NQL config corresponding to the query, NQL config_mode (correct answer) and NQL config_domain (movie domain) can be set.

In comparison with the query attributes illustrated in FIG. 5, NQL config_mode is the same as the correct answer, but NQL config_domain has a difference. As the NQL config_domain to be applied to the query changes, the type of data to be derived changes.

6, a poster picture 601, basic information 602, and additional information 603 for the movie are defined as basic exposure attributes for the movie query word, The details can be configured to be listed. At this time, NQL config_property can set the movie evaluation information (review) to be outputted with the highest priority when NQL config_domain is a movie domain.

5 and FIG. 6, when the search engine applies different NQL configs through analysis of the NQL attribute corresponding to the query language, the data type and data processing type provided according to the attribute of the query language Can be identified.

7 is a diagram illustrating another example of data laid out according to an NQL config reflecting a user query intention according to an embodiment of the present invention.

Referring to FIG. 7, when the user query word is "you (A drama) star (B drama)", the search engine according to the embodiment of the present invention searches the NQL attribute corresponding to the user query word Is a broadcasting program of the same kind, and the corresponding query type is "comparative type " and corresponds to" broadcasting domain ". Accordingly, NQL config_mode (comparison type) and NQL config_domain (broadcasting domain) can be set in setting NQL config corresponding to the query.

In other words, to compare two search objects as defined in NQL config_mode (comparison type), use search data type, maximum number of search data, number of attribute for search, number of attribute for result, etc., and use NQL config_domain ( Broadcasting domain), an image providing method, an alignment method, a link generating method, and the like can be used. In addition, NQL config_property can be used with defined conditions such as data type, data domain, OQL structure corresponding to the NQL, result data processing method, and basic exposure attribute to be derived for each attribute according to the setting.

7, data processing is performed in the form of comparing the useful information 701 to 706, which are useful for comparison in consideration of the fact that the two search targets are broadcast programs, in a bilaterally symmetrical manner, Can be laid out.

8 is a diagram illustrating another example of data laid out according to an NQL config reflecting a user query intention according to an embodiment of the present invention.

Referring to FIG. 8, when the user query term is "age of actors appearing in older sisters than flowers?", The natural language processing engine generates NQLs as illustrated in FIG. 8 (a) Can be generated. 801 "indicating a search target;" broadcast / player "802 indicating a search domain and an attribute;" person / age "803 indicating attribute information about a search target; And "person / jop" (804).

Next, the search engine according to the embodiment of the present invention can analyze that the query type corresponds to multiple domains such as " graph type " and "person domain, broadcast domain" Accordingly, in setting NQL config corresponding to the query, NQL config_mode (graph type) and NQL config_domain (person domain, broadcasting domain) can be set.

In the NQL config_mode (graph type), the search target entity and its surrounding data are provided together, and the search is performed to search the knowledge database mainly for attributes for calculation of intimacy between entities. Thus, as shown in FIG. 8B, Quot; person / job "entity 807 associated with the entity " sister" entity 805, a " broadcast / cast "entity 806, Level search to the " person / age "entity 808 associated with the " broadcast / cast" entity, and to perform a 2-2step search to the entity 805 to identify the relationship between the search entity 805 and its surrounding entities 806 to 808 The search result screen visualized by the graph of FIG.

The foregoing description is merely illustrative of the technical idea of the present invention, and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments of the present invention are not intended to limit the scope of the present invention but to limit the scope of the present invention. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents thereof should be construed as being included in the scope of the present invention.

Claims (14)

A mesh-up search language conversion step of converting the semantic-tagged natural language query into a mesh-up search language;
Up search environment corresponding to the converted mesh-up search language from the search environment information DB based on the attribute information of the mesh-up search language, and searches the converted mesh-up search language for the target search A target search language conversion step of converting the target language into a language;
A retrieval step of retrieving data corresponding to the converted target retrieval language from a knowledge database according to the retrieval environment information; And
And an output step of processing the retrieved data according to the retrieval environment information and laying out a retrieval result screen including the processed data,
The search environment information DB includes:
A search mode defined by a query type, a search domain defined by a search field included in a query, and a search property defined by a method of deriving a result value for each attribute, a search condition, and a search result provided for at least one of A method for extracting and providing information in accordance with a natural language query intention. The method according to claim 1,
Wherein the search mode defines at least one of a maximum number of search data, a number of attributes for search, and a number of attribute usage for each attribute according to a query type,
Wherein the search domain defines at least one of a basic exposure attribute, an image providing method, an alignment method, and a link generation method according to a domain attribute,
The search property includes at least one of a search result type, a result domain, a structure of a target search language corresponding to a mesh-up search language, a result data processing method, and a basic exposure attribute, and extracts and provides information corresponding to a natural language query intention Way. 3. The method according to claim 1 or 2,
Wherein the target search language conversion step comprises:
Analyzing attributes and domain information of the mesh-up search language;
Selecting search environment information corresponding to the mesh-up search language according to the analysis result;
Normalizing the mesh-up search language using the selected search environment information; And
And converting the normalized mesh-up search language into the target search language using the selected search environment information. 3. The method according to claim 1 or 2,
The retrieving step comprises:
Deriving one or more physical data corresponding to the converted target search language from the knowledge database according to the search environment information; And
And deriving result data from the derived one or more physical data using logical operation rules included in the search environment information. 3. The method according to claim 1 or 2,
If the search mode is the correct answer type,
Wherein the search environment information searches for one or more additional data associated with the basic data and the basic data corresponding to the user query and defines that the basic data and the one or more additional data are laid out on the same search result screen, A method for extracting and providing information in accordance with an intention. 3. The method according to claim 1 or 2,
If the search mode is a comparison type,
Wherein the search environment information includes data of the same attribute for each of the comparison objects included in the user query and defining that the attribute data for each of the searched comparison objects are laid out on the same search result screen, Information extraction and delivery method. 3. The method according to claim 1 or 2,
If the search mode is graphical,
Wherein the search environment information searches for one or more additional data related to the search object in accordance with the basic data for the search object corresponding to the user query and the query content, and the basic data and the one or more additional data are displayed on the search result screen And defining a layout in the form of a graph together with a relationship diagram. A system for extracting and providing information in accordance with a natural language query intention,
A natural language processing engine for normalizing a morpheme constituting a natural language query input through a user device and converting the natural language query into a mesh-up search language according to a dependency relationship between normalized morpheme parts;
Up search environment corresponding to the converted mesh-up search language is selected from the search environment information DB based on the attribute information of the mesh-up search language, and based on the selected search environment information, the converted mesh- A search engine for performing data search by converting the search data into a target search language and providing the search data according to the search environment information; And
And a knowledge database including at least one of the search environment information DB, the language conversion information DB, and the knowledge information DB,
The search environment information DB includes:
A search mode defined by a query type, a search domain defined by a search field included in a query, and a search property defined by a method of deriving a result value for each attribute, a search condition, and a search result provided for at least one of And a method for extracting and providing information that conforms to a natural language query intention. 9. The method of claim 8,
Wherein the search mode defines at least one of a maximum number of search data, a number of attributes for search, and a number of attribute usage for each attribute according to a query type,
Wherein the search domain defines at least one of a basic exposure attribute, an image providing method, an alignment method, and a link generation method according to a domain attribute,
The search property includes at least one of a search result type, a result domain, a structure of a target search language corresponding to a mesh-up search language, a result data processing method, and a basic exposure attribute, and extracts and provides information corresponding to a natural language query intention system. 10. The method according to claim 8 or 9,
The search engine comprises:
A mesh-up search language normalization module for analyzing attribute information of the mesh-up search language to select search-environment information corresponding to the mesh-up search language, and normalizing the mesh-up search language using the selected search environment information;
A target search language conversion module for converting the normalized mesh-up search language into the target search language according to the search environment information;
A data derivation module for retrieving and deriving physical data for the target search language from the knowledge database according to the search environment information to generate result data; And
And a data processing module for processing the result data generated by the data derivation module according to the search environment information and generating a search result screen including the processed data, . 11. The method of claim 10,
Wherein the data derivation module comprises:
Deriving one or more physical data corresponding to the converted target search language from the knowledge database in accordance with the search environment information, and using the logical operation rules included in the search environment information from the derived one or more physical data, And extracting and providing information in accordance with the natural language query intention. 11. The method of claim 10,
If the search mode is the correct answer type,
Wherein the data derivation module derives from the knowledge database basic data conforming to the user query and one or more additional data related to the basic data according to the search environment information,
Wherein the data processing module lays out a search result screen so that the basic data and the one or more additional data are included in the same screen in accordance with the search environment information. 11. The method of claim 10,
If the search mode is a comparison type,
Wherein the data derivation module derives data of the same attribute for each of the comparison objects included in the user query word from the knowledge database according to the search environment information,
Wherein the data processing module lays out a search result screen so that attribute data of each of the searched comparison objects is included in the same screen according to the search environment information. 11. The method of claim 10,
If the search mode is graphical,
Wherein the data derivation module derives one or more additional data related to the search object in accordance with the basic data for the search object corresponding to the user query term and the query content from the knowledge database according to the search environment information,
Wherein the data processing module lays out a search result screen such that the basic data and the one or more additional data are output in a graph form together with a relation on one screen of search results according to the search environment information, Delivery system.

Images