KR20120053253A - System and method for hybrid semantic searching service - Google Patents

Abstract

PURPOSE: A hybrid semantic search system and a method thereof are provided to improve search response speed and quality with respect to all kinds of user requests by selectively providing a semantic search process. CONSTITUTION: A query word receiver(110) receives a query word from a user terminal. An automatic sentence completion part(120) provides a completed sentence corresponding to the query word to the user terminal by searching a database(170) which includes completed sentences. An automatic sentence completion processing part(140) selectively provides a semantic search process based on properties of the completed sentence when the query word is selected from a list of the completed sentences.

Description

Hybrid semantic search system and method {SYSTEM AND METHOD FOR HYBRID SEMANTIC SEARCHING SERVICE}

The present invention relates to a system and method for providing a hybrid semantic search, and more particularly, to a system for providing a semantic search for performing a search by using a pre-stored information for a user's query and a search by query interpretation. It is about a method.

Due to the development of wired / wireless communication technology and the spread of the Internet, services using the Internet are expanding to various fields such as IPTV and Internet telephony. In addition, a variety of services for providing a search result for a user's query through the Internet have been diversified, and various methods have been applied to provide more accurate and rapid results for a user's query.

Among the conventionally provided search methods, the most commonly used keyword search method is performed by comparing simple strings and keywords, and this keyword search method can find a result relatively quickly by using keyword indexing in file form, There was a problem that the search was not possible if the matched words or phrases were not included.

Semantic search, which is used to overcome the shortcomings of such a simple keyword indexing scheme, is gradually provided.

Semantic search is a search method that analyzes the user's query words, grasps the intentions, and semantically finds the user's target through semantic reasoning. Search based on actual content, not search.

In this case, in semantic search, the objects constituting the search object as components according to the search domain, the objects constituting the search condition have various characteristics and forms, and the user's query has a simple keyword form directly referring to the object. It may be in the form of a natural language sentence describing the characteristics of the object.

Therefore, in the conventional semantic search, most of all query requests first perform a query analysis to determine the intention, and then perform a direct search in the ontology database based on the query analysis result to provide a result.

In this case, according to the conventional semantic search, since the stochastic complex analysis process must be performed using a system library trained using a statistical method, natural language interpretation requires considerable resources and time, and the structure of the ontology is complicated and large. If you have to go through a complicated reasoning process, it may take a long search time.

However, the actual commercial search service must respond quickly and quickly to a search request required by a large group of customers at the same time, and satisfy the semantic-based intelligent search and response time performance simultaneously. There is a problem in that an application service using a semantic search service cannot be provided smoothly.

An embodiment of the present invention does not use a simple keyword matching method or a batch semantic search method, but selectively uses a keyword indexing search, a search based on a pre-stored query, and a search based on a query interpretation according to a user's query. An object of the present invention is to provide a system and method for efficiently providing optimized search results.

As a technical means for achieving the above-described technical problem, the first aspect of the present invention is (a) determining the type of query received from the user terminal, (b) when the query is a free query input by the user Performing a keyword search for the query, (c) determining whether to search for the semantic based on the query based on the keyword search result of the query, and ontology on the query based on the determination. A method of providing a semantic search by a hybrid semantic search system comprising: performing a semantic based search using the method and (d) providing the user terminal with the semantic based search result performed in the step (c). Can be.

In a first aspect of the present invention, in the step (c), when the keyword search result for the query word does not correspond to a preset criterion, a semantic based search using ontology may be performed on the query word. .

In the first aspect of the present invention, in the step (c), if the number of keyword search results for the query is less than a predetermined reference value, it may be determined that the keyword search results do not correspond to the preset criteria.

In the first aspect of the present invention, the step (c) may include (c1) a semantic based search query that is preset to match the semantic query when a pre-stored semantic query that matches the query exists in an autocomplete sentence database. The method may include performing a semantic based search using an ontology for the query based on the term.

Also, in the first aspect of the present invention, the step (c) may include (c2) if the result of the keyword search and the semantic based search based on the preset semantic based search query does not correspond to a preset criterion, Parsing a query to extract a named entity (NE) from the query, (c3) generating a query for meaning-based search based on the extracted entity name (NE), and (c4) The method may include performing a semantic based search using an ontology for the query based on a semantic based search query.

Also, in the first aspect of the present invention, the step (c) may include (c5) if the pre-stored semantic query word matching the query word does not exist in the autocomplete sentence database, interpreting the query word to determine the entity name from the query word. extracting a named entity (NE), (c6) generating a semantic based search query based on the extracted entity name (NE), and (c7) based on the generated semantic based search query It may include performing a semantic based search using an ontology for the query.

A second aspect of the present invention is a query input receiving unit for receiving a query from a user terminal, an automatic completion sentence for searching for a list of autocomplete sentences matching the query received from the user terminal from an autocomplete sentence database and providing the query to the user terminal. If the query unit received from the user is selected from a list of the autocomplete sentences provided by the autocomplete sentence providing unit, based on the attributes of the autocomplete sentences, keyword search or semantic based on the autocomplete sentences An automatic completion sentence processing unit for selectively performing a search, and if a query received from the user terminal is not selected from a list of the automatic completion sentences provided by the automatic completion sentence providing unit, search for a keyword based on a search result for the query; And ont A free query processing unit that sequentially performs a semantic based search using an ontology, and a response providing unit providing the user terminal with a search result performed on the query by the autocomplete sentence processing unit or the free query processing unit; A hybrid semantic search system can be provided.

In the second aspect of the present invention, the autocomplete sentence processing unit may perform keyword search on the autocomplete sentence indicating a specific name.

In addition, in the second aspect of the present invention, the autocomplete sentence processing unit ontology for the autocomplete sentence based on the semantic based search query that is stored matched with the autocomplete sentence for the autocomplete sentence that requires a meaning interpretation A semantic based search can be performed using.

According to the above-described problem solving means of the present invention, the present invention does not perform a lot of resources and a long time interpretation and inference process for all query words, intelligently select a search method according to the characteristics of the input query keyword indexing By providing search, semantic search based on pre-stored SparQL query and semantic search by query interpretation, search response speed and quality can be improved for all types of user queries.

1 is a diagram illustrating a configuration of a hybrid semantic search system according to an embodiment of the present invention;
2 is a diagram showing the configuration of a free query processing unit according to an embodiment of the present invention;
3 is a flowchart illustrating a method of performing a search on an autocomplete sentence input by a user in an autocomplete sentence processing unit according to an embodiment of the present invention;
4 is a flowchart illustrating a method of performing a search for an autocomplete sentence input by a user by a free query processing unit according to an exemplary embodiment of the present invention.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and like reference numerals designate like parts throughout the specification.

Throughout the specification, when a part is referred to as being "connected" to another part, it includes not only "directly connected" but also "electrically connected" with another part in between . In addition, when a part is said to "include" a certain component, which means that it may further include other components, except to exclude other components unless otherwise stated.

1 is a diagram illustrating a configuration of a hybrid semantic search system according to an embodiment of the present invention.

The hybrid semantic search system 100 according to an embodiment of the present invention includes a query input receiver 110, an autocomplete sentence providing unit 120, a query type analysis unit 130, an autocomplete sentence processing unit 140, and a free query processing unit. 150 and the response generator 160.

The query input receiving unit 110 receives a query input by the user from the user terminal. The user terminal includes a terminal capable of wired and wireless communication such as a desktop, a notebook computer, a smartphone, a personal digital assistant (PDA), and the like.

The user can search for a word or sentence in the form of a word or sentence, such as the title of the content to be searched through a query input means, for example, 'Good Morning President', 'Cinderella', 'Kang Woo-seok', 'Infinite Challenge Kim Yu-na', You can enter the query, such as "Let's play."

In addition, the query input receiving unit 110 receives not only a query that has been input by the user but also a query that is being input. For example, when a user inputs a 'good morning presenter', the query input receiving unit 110 may query 'a', 'ㄱ TT', 'good', .... 'Mo', 'Ning', 'F', 'Le', 'G', 'Dun', 'T' can be received sequentially, and the initial consonant of the search query that the user wants to query, such as 'ㄱ ㅁㄴㄴ ㄹㅈ ㄷㅌ' It is also possible to receive input words consisting of bays.

The autocomplete sentence providing unit 120 analyzes the query received by the query input receiving unit 110, searches for the autocomplete sentence stored in the autocomplete sentence database 170 based on the analysis result, and provides the same to the user.

That is, the autocomplete sentence providing unit 120 is an autocomplete sentence including the word or phoneme of the query word input by the user so far, that is, a sentence stored in the autocomplete sentence database 170 in advance in a frequently used query word. To the user.

The autocomplete sentence providing unit 120 may search for an autocomplete sentence from pre-established indexing information based on a combination of phoneme units or syllable units including a continuous initial input form, and the indexing information may be an autocomplete sentence. And a Named Entity (NE) based actual SparQL query that matches the autocomplete statement.

A user may request a search by selecting an autocomplete sentence provided by the autocomplete sentence providing unit 120 or directly input a query sentence without selecting the autocomplete sentence and request a search.

The query type analysis unit 130 analyzes the type of the query input by the user to determine whether to process the query in the automatic completion sentence processing unit 140 or the free query processing unit 150.

The query input by the user may include an autocomplete sentence provided by the autocomplete sentence providing unit 120 and a free query word freely input by the user. The autocomplete sentence may include a simple name-based query and a compound semantic query. have.

The query type analysis unit 130 obtains an entity name (NE) based SpaQL query that matches the autocomplete sentence from the autocomplete sentence database 170 when the input query is an autocomplete sentence, and then completes the autocomplete sentence. And transmits the obtained entity name (NE) based SpaQL query to the autocomplete sentence processing unit 140.

In addition, when the input query is a free query, the query type analysis unit 130 transmits the free query to the free query processing unit 150.

The autocomplete sentence processing unit 140 searches for a keyword based on the auto complete sentence received from the query type analysis unit 130 and the entity name (NE) SparQL query, or the entity name (NE) based SpaQEL ( SparQL) query statement search, and provides the search results to the response generator 160.

That is, when the attribute of the autocomplete sentence received from the query type analysis unit 130 corresponds to a simple name indicating a specific object, the autocomplete sentence processing unit 140 generates a keyword for the word included in the received autocomplete sentence. Perform a search.

In addition, when the attribute of the autocomplete sentence received from the query type analysis unit 130 corresponds to a compound semantic query word requiring semantic interpretation, the autocomplete sentence processing unit 140 stores the entity name previously stored for the received autocomplete sentence. (NE) -based SparQL query is acquired and semantic-based search is performed using the obtained SparQL query.

The free query processing unit 150 performs a keyword search, a semantic based search based on an autocomplete sentence, and a semantic based search based on the query analysis, on the free query received from the query type analysis unit 130, and generates a search result. Provided at 160.

The response generator 160 generates a response to the query input by the user using a search result received from the autocomplete sentence processor 140 or the free query processor 150 and provides the generated response to the user. .

2 is a diagram illustrating a configuration of a free query processing unit according to an embodiment of the present invention.

The free query processing unit 150 according to an embodiment of the present invention includes a keyword indexing search unit 152, an autocomplete sentence comparison processing unit 154, a query analysis processing unit 156, a semantic ontology database 157, and SparQL. ) Query database 158.

The keyword indexing searcher 152 performs a keyword indexing search on the keywords included in the free query using a pre-stored index file for the free query input by the user.

The keyword indexing search unit 152 compares the keyword indexing search result with a preset reference value and determines whether to provide only the keyword indexing search result to the user or perform an additional search.

The autocompletion sentence comparison processing unit 154 obtains an autocompletion sentence from an autocompletion sentence database (not shown) that can be provided for a free query word whose keyword indexing search result does not satisfy a predetermined criterion.

The autocompletion sentence comparison processing unit 154 obtains a compound semantic based semantic query from the obtained autocompletion sentences, and matches a SPARQL query matching the semantic query obtained from the SparQL query database 158. Obtain a door.

The automatic completion sentence comparison processing unit 154 performs a semantic based search using the semantic ontology database 157 based on the obtained SparQL query.

The autocompletion sentence comparison processing unit 154 may search the keyword indexing search result by the keyword indexing search unit 152 and the search result using the pre-stored SparQL query statement by the autocompletion sentence comparison processing unit 154 and the preset criteria. Compare the values to determine whether to search further.

When the query analyzing processing unit 156 requires additional search other than the search using the keyword indexing search by the keyword indexing searching unit 152 and the pre-stored SparQ query by the automatic completion sentence comparing processing unit 154. After analyzing the query input by the user, a named entity (NE) is obtained, and a SparQL query is generated using the obtained entity name (NE).

The query analysis processing unit 156 performs a semantic based search using the semantic ontology database 157 based on the generated SparQL query and provides the search result to the user.

3 is a flowchart illustrating a method of performing a search on an autocomplete sentence input by a user according to an embodiment of the present invention.

In step S110, it is determined that the query input by the user by the query type analysis unit (not shown), that is, the query received from the user terminal is an autocomplete sentence, and in step S120, the autocomplete sentence processing unit is a query. Receives an autocomplete sentence selected by the user from a list of autocompletion sentences provided by the user, that is, a query input by the user from the type analyzer.

The autocomplete sentence processing unit may receive a SpaQL query matching the autocomplete sentence together with the autocomplete sentence.

In step S130, the autocompletion sentence processing unit analyzes the type of the autocompletion statement, that is, attributes, to search whether the autocompletion sentence received in step S120 corresponds to a simple query, for example, a simple name or semantically. Determines whether it corresponds to a compound semantic query.

In step S140, when it is determined in step S130 that the attribute of the autocomplete sentence corresponds to a simple query, the autocomplete sentence directly refers to the name of the search target, and thus is stored in advance to retrieve such a name quickly. Perform a keyword indexing search to search the index file.

For example, if the autocompletion sentence corresponds to a simple name of a service domain that is targeted such as 'good morning president' or 'Cinderella', for example, the title of the video on demand (VOD) content to be searched, the autocompletion sentence processing unit A search may be performed using a pre-stored index file for 'Good Morning President' or 'Cinderella'.

In step S150, when it is determined in step S130 that the attribute of the autocomplete sentence corresponds to a complex semantic query term instead of a simple query term, the autocomplete sentence is, for example, "TV entertainment of Jang Dong Gun". If it is judged to include both the simple name and its meaning-based condition, or if the autocomplete sentence is judged to describe only the meaning of the object to be searched for, such as "Action Movie directed by Kang Woo-seok," the autocomplete sentence processing unit In operation S120, a SPARQL query that matches an autocomplete sentence is searched for and obtained from an autocomplete sentence database (not shown).

When the autocomplete sentence processing unit receives a SpaQL query matching the autocomplete sentence received together with the autocomplete sentence in step S120, the search for the SpaQL query may be omitted.

In step S160, the autocompletion sentence processing unit uses the SpaQL query obtained in step S150 or the SpaQL query received with the autocompletion sentence in step S120 to mean from the ontology. Performs a base search to search for autocomplete sentences entered by the user.

As such, in one embodiment of the present invention, when the user selects one autocomplete sentence from one or more autocomplete sentences provided by the hybrid semantic search system, the hybrid semantic search system according to an embodiment of the present invention is an autocomplete sentence. The search result can be provided to the user more efficiently by performing a search according to the type of.

FIG. 4 is a flowchart illustrating a method of performing a search for an autocomplete sentence input by a user by a free query processing unit according to an exemplary embodiment of the present invention.

In step S205, it is determined that the query input by the user by the query type analysis unit (not shown), that is, the query received from the user terminal is a free query, and in step S210, the free query processing unit analyzes the query type. The query receives the query input by the user, that is, the query directly input by the user without selecting the provided autocomplete sentence.

In step S215, the free query processing unit performs a keyword indexing search on the query received in step S210 by using a pre-stored index file. That is, the free query processing unit assumes that the query input by the user is a query based on a general name before performing a query analysis on the query, and performs a keyword indexing search on the query.

In step S220, the free query processing unit determines whether the number of results searched by the keyword indexing search performed in step S215 is greater than or equal to a predetermined reference value, that is, whether the keyword indexing search alone can obtain sufficient search results. .

The results of the search performed based on the match of the query and the general name correspond to the higher priority among the search rankings. Therefore, if the number of keyword search results is higher than or equal to the preset reference value, the higher priority search results are sufficient. Can provide. Therefore, the number of keyword indexing search results is compared with a preset reference value.

In step S225, when it is determined in step S220 that the number of keyword indexing search results is less than a predetermined reference value, the free query search unit may determine an autocomplete sentence based on the query received in step S210 from the autocomplete sentence database. Search.

That is, the free query search unit searches for an autocomplete sentence to be provided by an autocomplete sentence providing unit (not shown) when the user inputs the query received in operation S210 from the autocomplete sentence database.

In step S230, the free query processing unit determines whether a complex semantic based semantic query is included in the one or more autocomplete sentences retrieved in step S225. In other words, the free query processing unit searches for a complex semantic based semantic query among the one or more autocomplete sentences found in step S225.

In step S235, when it is determined in step S230 that the one or more autocomplete sentences searched on the basis of the query includes a complex semantic based semantic query, the free query processing unit is included in the one or more autocomplete sentences searched on the basis of the query. A prestored SparQL query that matches the semantic query included is retrieved from an autocomplete sentence database.

In step S240, the free query processing unit performs a semantic based search from the ontology using the SparQL query obtained in step S235. In this case, since the free query processing unit performs a predefined semantic query search using a pre-stored SparQL query, it is possible to perform the search more efficiently and quickly.

In step S245, the free query processing unit adds the sum of the number of keyword indexing search results performed in step S215 and the number of semantic based search results using a pre-stored SparQ query performed in step S240. It is determined whether the reference value is greater than or equal to a preset reference value.

In step S250, when it is determined in step S245 that the sum of the number of keyword indexing search results and the number of semantic based search results using a pre-stored SparQL query is less than a predetermined reference value, the free query processing unit Determines that the search result does not correspond to the level to be provided to the user, and interprets the free query to determine the meaning of the free query input by the user received in step S205.

That is, the free query processing unit analyzes the free query to extract semantic entities (Named Entity, NE) included in the free query in order to perform semantic search using the ontology for the user's free query.

In step S255, the free query processing unit generates a SpaQL query using the entity name NE extracted in step S250.

For example, when the free query input by the user is a movie starring Jang Dong-gun and Ko So-young, the free query processing unit may refer to the name of the object corresponding to 'actor' (Jang Dong-gun) and 'Go So-young' in step S250 ( NE), 'appearance' can be extracted as an entity name (NE) corresponding to 'action', and 'movie' can be extracted as an entity name (NE) corresponding to 'target'.

Therefore, the free query processing unit uses the individual names (NE) corresponding to 'Jang Dong Gun', 'Go So Young', 'Appearance', and 'Movie' extracted from the free query 'Jang Dong Gun and So So Young's movie'. (SparQL) You can create a query.

In step S260, the free query processing unit performs a semantic based search using an ontology using the SparQL query generated in step S255.

In step S265, the free query processing unit provides a keyword indexing search result performed in step S200 to the user according to the determination in step S215, step S230, and step S245, or step S200. Spam generated by analyzing a user's query or providing a search result including a keyword indexing search result performed in the) and a semantic based search result based on a pre-stored SparQL query executed in step S240. A semantic based search result based on a SparQL query can be provided.

As described above, according to an exemplary embodiment of the present invention, a search method is performed by applying a search method that is easy to search and has a fast search speed to a query input by a user, and performs a search according to a search result. If not, the user can apply the semantic based search by the pre-stored SparQL query and the semantic based search by the SparQL query generated by the query interpretation. It can provide a search response result that satisfies.

One embodiment of the present invention may also be embodied in the form of a recording medium including instructions executable by a computer, such as program modules, being executed by a computer. Computer readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media.

In addition, computer readable media may include both computer storage media and communication media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Communication media typically includes computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave, or other transmission mechanism, and includes any information delivery media.

While the methods and systems of the present invention have been described in connection with specific embodiments, some or all of those elements or operations may be implemented using a computer system having a general purpose hardware architecture.

The foregoing description of the present invention is intended for illustration, and it will be understood by those skilled in the art that the present invention may be easily modified in other specific forms without changing the technical spirit or essential features of the present invention. will be. It is therefore to be understood that the above-described embodiments are illustrative in all aspects and not restrictive. For example, each component described as a single entity may be distributed and implemented, and components described as being distributed may also be implemented in a combined form.

The scope of the present invention is defined by the appended claims rather than the detailed description and all changes or modifications derived from the meaning and scope of the claims and their equivalents are to be construed as being included within the scope of the present invention do.

100: Hybrid Semantic Search System
110: query input unit 120: automatic completion sentence providing unit
130: query type analysis unit 140: automatic completion sentence processing unit
150: free query processing unit 160: response generation unit

Claims (16)

In the method by which the hybrid semantic search system provides semantic search,
(a) determining the type of query word received from the user terminal;
(b) if the query is a free query input by a user, performing a keyword search on the query,
(c) determining whether or not a semantic based search is performed for the query based on a keyword search result for the query, and performing a semantic based search using ontology on the query based on the determination;
(d) providing the user terminal with the semantic based search result performed in step (c)
Semantic search providing method comprising a.
The method of claim 1,
In step (c), if the keyword search result for the query does not correspond to a preset criterion, the semantic search providing method of performing a semantic based search using an ontology on the query.
The method of claim 2,
In the step (c), if the number of keyword search results for the query is less than a predetermined reference value, it is determined that the keyword search results do not correspond to a predetermined criterion.
The method of claim 2,
In step (c),
(c1) when a pre-stored semantic query word matching the query word exists in an autocomplete sentence database, a semantic based search using an ontology of the query word is performed based on a semantic-based search query word which is preset to match the semantic query word. Steps to perform
How to provide a semantic search.
The method of claim 4, wherein
In step (c),
(c2) when the result of the keyword search and the semantic-based search based on the preset semantic-based search query does not correspond to a preset criterion, the query is interpreted to obtain a named entity (NE) from the query. Extracting step,
(c3) generating a query for meaning-based search based on the extracted entity name (NE); and
(c4) performing a semantic based search using an ontology for the query based on the generated semantic based search query
The semantic search providing method further comprising.
The method of claim 2,
In step (c),
(c5) if no pre-stored semantic query word matching the query word exists in the autocomplete sentence database, extracting a named entity (NE) from the query word by analyzing the query word;
(c6) generating a query for meaning-based search based on the extracted entity name (NE); and
(c7) performing a semantic based search using an ontology for the query based on the generated semantic based search query
How to provide a semantic search.
The method of claim 1,
The step (a)
(a1) sequentially receiving components of the query word from the user terminal;
(a2) retrieving an autocomplete sentence matching the sequentially received query component from a prestored autocomplete sentence database; and
(a3) transmitting an autocomplete sentence list including the found one or more autocomplete sentences to the user terminal
How to provide a semantic search.
The method of claim 7, wherein
(e) If the query is an autocomplete sentence included in the list of autocomplete sentences transmitted to the user terminal in (a3), keyword search or semantic based on the autocomplete sentence based on the attribute of the autocomplete sentence Perform a search, and
(f) providing the user terminal with the search results performed in step (e)
The semantic search providing method further comprising.
The method of claim 8,
In step (e),
(e1) performing a keyword search based on the name when the autocomplete sentence indicates a name of an object indicated by a word included in the autocomplete sentence;
How to provide a semantic search.
The method of claim 8,
In step (e),
(e2) when the autocomplete sentence corresponds to a semantic query, a semantic based search using an ontology of the autocomplete sentence is performed based on a preset semantic based search query matched with the autocomplete sentence. Steps to
How to provide a semantic search.
In the hybrid semantic search system,
Query input receiver for receiving a query from the user terminal,
An autocomplete sentence providing unit for searching for a list of autocomplete sentences matching the query received from the user terminal from an autocomplete sentence database and providing the list to the user terminal;
If the query received from the user is selected from a list of autocomplete sentences provided by the autocomplete sentence providing unit, select a keyword search or a semantic based search for the autocomplete sentences based on the attributes of the autocomplete sentences. Autocomplete sentence processing unit to perform,
If the query received from the user terminal is not selected from the list of the autocomplete sentences provided by the autocomplete sentence providing unit, the keyword search and the semantic based search using ontology are performed based on the search result for the query. Free query processing unit to perform sequentially and
A response providing unit providing the user terminal with a search result performed on the query by the automatic completion sentence processing unit or the free query processing unit;
Semantic search system comprising a.
The method of claim 11,
The free query processing unit
A keyword indexing search unit that performs a keyword search based on the words included in the query word,
When the number of keyword search results by the keyword indexing search unit is less than a preset reference value, performing a semantic based search using an ontology of the query based on the semantic based search query matched with the query and stored; A semantic search
When there is no semantic-based search query stored in accordance with the query, the semantic-based search query is generated by analyzing the query, and an ontology for the query is based on the generated semantic-based search query. Semantic based retrieval unit performing semantic based retrieval using
The semantic search system that includes.
The method of claim 12,
The first semantic based searcher obtains an autocomplete sentence matching the query from the autocomplete sentence database, and obtains a semantic based search query matching the obtained autocomplete sentence, thereby searching for the semantic based on the query. To perform the semantic search system.
The method of claim 12,
And the second semantic based searcher extracts a named entity (NE) from the query and generates the semantic search query based on the extracted entity name (NE).
The method of claim 11,
And the automatic completion sentence processing unit performs a keyword search on the automatic completion sentence indicating a specific name.
The method of claim 11,
The autocomplete sentence processing unit performs a semantic based search using an ontology for the autocomplete sentence based on a semantic based search query sentence matched with the autocomplete sentence and stored for the autocomplete sentence requiring semantic interpretation. Search system.

Images