KR20100039968A - Ontology based semantic search system and method for authority heading of various languages via automatic language translation - Google Patents

Abstract

PURPOSE: An ontology based semantic search system for authority headings of various languages for automatic language translation and a method thereof are provided to perform semantic search in a resource database constructed by ontology based on topic map structure. CONSTITUTION: Multiple authority heading extractors(151,161,171) decide whether there is an authority heading corresponding to a keyword of a corresponding language. An authority heading extraction units extract the authority heading. About the keyword translated into the language and the authority heading, multiple search units(153,163,173) perform search in a database based on the language among multiple resource databases(154,164,174). By collecting search results of respective languages, a search result providing unit(180) provides summary information according to a pre-designated display order.

Description

Ontology Based Semantic Search System and Method for Authority Heading of Various Languages via Automatic Language Translation}

The present invention relates to an ontology-based semantic search system and method, and more particularly, to automatically input into a multi-language input search word so that a rich search result for various languages can be realized regardless of language, The present invention relates to an ontology-based semantic retrieval system and method that can perform accurate retrieval by providing accurate ontology access by extracting authority headings for language.

In today's flood of information, a system that can quickly and accurately search for information that a user wants to find is required in many fields. For example, a lot of information managed by a system that is useful for searching, such as a book search system in a library, an information search system on a web site, a search system for resources managed on a small network connected to a LAN or a storage device of a personal PC. For example, many methods for retrieving various types of information, such as document files, image files, video files, database information, and text data, in a faster and more intelligent manner, have been developed.

In general, a search is performed in a language of a country, and in the search, an approach of a keyword extraction through morphological analysis or an n-gram method through an index structure is used to overcome limitations of a morphological dictionary. Such conventional search methods only provide a search result by extracting a string corresponding to a search word from a search target data and comparing the matched word with the search word. However, these search methods do not provide data in various national languages related to the search terms, and rely on simple form analysis in the search approach, and thus tend to include a large amount of data that is inappropriate for search terms as a search result.

In order to overcome this problem, the conventional search method may be supplemented by extracting a similar word for a search word from a thesaurus dictionary and referring to the degree of correlation when comparing the match. However, even if you search by referring to the thesaurus, it will only provide search results for various languages, and the search will be started and performed only for the entered search terms, so that you can obtain satisfactory or various and accurate results according to the user's intention. There is a limit to expectation. As an alternative to this limitation, we will manage authority headings that the thesaurus does not provide and propose an ontology-based semantic search method.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and an object of the present invention is to provide an ontology-based meaning that can perform a semantic search on a resource database constructed with an ontology based on a topic map structure in order to provide rich and accurate search results. To provide a search system and method.

In addition, another object of the present invention, in order to realize a meaning search, even if a search word is input in any language is automatically translated into a multi-language to start the search in the language of each country, at this time, to translate more accurate for the user's intention To provide an ontology-based semantic search system and method that extracts authority headings for each language of each country, launches and aggregates the search for each authority heading, and provides the corresponding search results with high accuracy or frequency for each authority heading. have.

In addition, another object of the present invention, the display order of the search results can be determined by giving a weight to each search term in the display window of the search screen that can be searched given a plurality of combinations of search terms, in the search results screen The present invention provides an ontology-based semantic retrieval system and method that can provide information by dividing depth of information.

First, to summarize the features of the present invention, the ontology-based semantic retrieval system according to an aspect of the present invention for achieving the above object, a plurality of search terms translated in each country language of a plurality of countries with respect to the input search word Multilingual translator to generate; A plurality of language recognition units for recognizing and receiving a search word in each language of each of the translated plurality of search words; A plurality of authority heading databases for storing and managing authority headings for each word type for a plurality of words in each country language; A plurality of resource databases, each storing and managing resources constructed ontologies based on a topic map structure according to each country language; With respect to a search word in each country language from the plurality of language recognition units, whether there is an authority head corresponding to the search word in each country language by referring to a database based on the language of each country among the plurality of authority heading databases. A plurality of authority head extractors configured to determine and extract the authority head; A plurality of search units that perform a search in a database based on each country language of the plurality of resource databases for a search word translated into a corresponding language and authority heads in each country language from the plurality of authority head extractors; And a search result providing unit which collects search results for each language extracted by performing a search and provides summary information in a predetermined display order.

Each of the plurality of search units performs a search one by one according to the search rank by determining a search rank based on a sequence of characters of the corresponding language for a search word translated in the corresponding language and a plurality of authority headings in the corresponding language. When the search is performed, PSIs having a frequency of at least one search term or authority head are extracted from resources including each search term or authority head.

The plurality of authority heading databases are characterized in that the authority headings are databased according to word forms including personal names, place names, other proper nouns, common nouns, adjectives / adverbs, or verbs.

The ontology-based semantic search system further includes a user interface for providing a search term input window to input a logical combination (including AND or OR) of a plurality of search terms on a screen, and the multilingual translator performs translation for all of the plurality of search terms. And generate a plurality of search words translated in each country language for each search word, wherein the plurality of search parts correspond to the plurality of search words translated in a corresponding language and the plurality of search words from the plurality of authority head extractors, respectively. For each authorized heading, a search may be performed in a database based on the language of each of the plurality of resource databases.

The plurality of search units respectively search the plurality of search terms translated in the corresponding language and a plurality of authority headings in the corresponding language corresponding to each of the plurality of search terms, based on the order of character arrangement of the corresponding language. A frequency of each of the search terms or authority headings while satisfying the logical combinations among resources including each search term or authority headings for a plurality of search terms inputted to the logical combinations when performing the search. PSI may be extracted at least once.

The user interface may provide a weight setting window for the plurality of search terms, and the search result providing unit may control the user interface to display search results on the screen according to the weight order set in the weight setting window.

The user interface may provide a viewing step setting window on a search result screen, and the search result providing unit may control the user interface to classify and display the depth of the search result according to the viewing step set in the viewing step setting window.

The plurality of resource databases are distributed large capacity for document files, image files, video files, database information, or text data in a small network or personal PC connected to a library book retrieval system, website information retrieval system, LAN Information can be linked to a map of semantic relationships around topics to provide resources in conjunction with storage means that store ontologies.

The ontology-based semantic retrieval method according to another aspect of the present invention comprises the steps of: database a authority heading for each word type for a plurality of words in each language; Constructing and databasening resources as ontology based on a topic map structure according to each language; Generating a plurality of search words translated in each country language of the plurality of countries with respect to the input search word; Recognizing and receiving a search word in each language among the translated plurality of search words; Determining whether there is an authority heading corresponding to the search term in each country language by extracting the authority heading with reference to the database of authority headings for the search term in each country language; Performing a search on the database resources for a search term translated in a corresponding language and authority headings in each language; And collecting search results for each language extracted by performing the search and providing summary information in a predetermined display order.

According to the ontology-based semantic search system and method according to the present invention, a rich and accurate search result can be provided by performing a semantic search on a resource database constructed with an ontology based on a topic map structure.

In addition, according to the ontology-based semantic search system and method according to the present invention, even if a search word in any language is automatically translated into a multi-language to start the search in the language of each country, the translated country for more accurate search for the user's intention The semantic search can be realized by extracting authority headings for the language, starting and collecting the search for each authority heading, and providing the corresponding search results with high agreement or frequency for each authority heading.

In addition, according to the ontology-based semantic search system and method according to the present invention, the display rank of the search results can be determined by giving a weight to each search word in a display window of a search screen that can be searched by giving a search condition in multiple combinations of search terms. In the search result screen, the depth of information may be classified and provided for each viewing step.

In order to fully understand the present invention, the operational advantages of the present invention, and the objects achieved by the practice of the present invention, reference should be made to the accompanying drawings which illustrate preferred embodiments of the present invention and the contents described in the accompanying drawings.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram illustrating an ontology-based semantic search system 100 according to an embodiment of the present invention. Referring to FIG. 1, the ontology-based semantic retrieval system 100 according to an embodiment of the present invention may include a multilingual translator 110, a plurality of language recognition units 120, 130, and 140, and a plurality of search engines 150. , 160, 170, a plurality of authority heading databases 152, 162, 172, a plurality of resource databases 154, 164, 174, and a search result provider 180, and a plurality of search engines ( 150, 160, and 170 are authorized head extracting units 151/161/171 for extracting authority heads for a corresponding language among the plurality of authorized head extracting units 151, 161, and 171, respectively, and a plurality of provided PSIs. The PSI search unit 153/163/173 performs a search for a corresponding language among the search units 153, 163, and 173. Here, a plurality of language recognizers 120, 130, 140, a plurality of search engines 150, 160, 170, a plurality of authority heading databases 152, 162, 172, and a plurality of resource databases 154 , 164, and 174 are shown as being composed of three sets each to perform a semantic search for Korean, English, and Japanese, but the present invention is not limited thereto, and similar sets of components may be added for similar processing for more languages. It may be provided. The overall components of the ontology-based semantic retrieval system 100 may be realized by software, hardware, or a combination thereof that can be recognized by a computer.

The operation of the ontology-based semantic search system 100 will be described with reference to the conceptual diagram of FIG. 2.

The multi-language translator 110 generates a plurality of search words translated in each country language of the plurality of countries with respect to the input search word. According to the operation of the user interface, a search term input window may be provided on a computer screen such as a personal computer, and the user may input a search term in a search term input window provided on the screen (S210). Accordingly, before the search engines 150, 160, and 170 start authority search for each language, the multilingual translator 110 constructs the inputted search word into Korean, English, Japanese, Chinese, Spanish, .. According to the system environment has been translated into each language, and outputs a plurality of search terms generated by the translation (S220). Here, the multilingual translator 110 may include a configuration for translating to the same language as the search word input by the user. Accordingly, a word or phrase that is substantially the same as the search term input by the user is also included with the search terms translated into another language. It may be output as a translated search term. At this time, a database of dictionaries for each country may be used to perform the translation.

The plurality of language recognition units 120, 130, and 140 may each recognize and receive a search word in a corresponding language among the plurality of search words translated by the multilingual translator 110 (S230). For example, the Korean recognizer 120 may recognize and accept a Korean search word among a plurality of search words translated by the multilingual translator 110. Also, the English recognizer 120 may search for an English search term among the plurality of search words translated by the multilingual translator 110, and the Japanese recognizer 120 may search for among the plurality of search words translated by the multilingual translator 110. Recognize and accept Japanese search terms.

The plurality of authority head extracting units 151, 161, and 171 may include a plurality of authority head databases (152/162/172), respectively, for a search word in a corresponding language from the plurality of language recognition units 120/130/140. By referring to the database based on the language, it is possible to determine whether there is an authority heading corresponding to a search word in each language and extract the authority heading (S240). For example, the Korean authority heading extractor 151 refers to the Korean authority heading database 152 for a search word in Korean from the Korean recognition unit 120, and whether there is an authority head corresponding to the corresponding search word in Korean. Can be determined and the authorized heading can be extracted. The English authority head extractor 161 refers to the English authority heading database 162 with respect to English words from the English recognition unit 130 and determines whether there is an authority head corresponding to the corresponding English search word. Extracting the authority heading, and the Japanese authority heading extraction unit 171 refers to the Japanese authority heading database 172 with respect to the Japanese search word from the Japanese recognition unit 140, and the authority corresponding to the corresponding Japanese search word. The presence of a heading can be determined and the authority heading can be extracted.

Each of the authority heading databases 152, 162, and 172 stores and manages authority heading for each word type for a plurality of words in a corresponding language. For example, as illustrated in FIG. 3, 'Lonardo', which may be referred to as the same person as 'DaVinci', may be managed as an authority heading for the searched word “DaVinci” translated into Korean. Similarly, in terms of the words 'Da Vinci', 'ダ ヴ ィ ン チ', '? 번역' translated in the languages of the multilingual translator 110, 'Leonardo', 'レ オ ナ ル ド', '雷? 那 多' can be managed as an authority heading in each corresponding database.

In addition, as an example of other authority headings for Korean, 'Lee Kwang-soo' can be matched with 'Chunwon' and 'KBS' can be matched with 'Korea Broadcasting Corporation'. These headings are distinct from thesaurus and include names (eg, Kim Hong-do: unit), names (eg, Seoul: Hanyang), other proper nouns (eg, Yangtze: Yangzigang), common nouns. Depending on each word type, including for example Korean (Korean), adjectives / adverbs (for example, cool: excellent), or verbs (for example, walk: save) In the present invention, each authority heading is a new search word, and a search is started for each heading to extract a search result. The existing method of searching by referring to thesaurus dictionary does not start a separate search for each thesaurus word, but only utilizes it in determining the degree of correlation when comparing the match with the resource to be searched.

Meanwhile, each of the plurality of PSI search units 153, 163, and 173 includes a plurality of resources for a search word translated in a corresponding language and authority heads in a corresponding language from the plurality of authority head extracting units 151, 161, and 171. A search is performed in a database based on the corresponding language among the databases 154, 164, and 174 (S250). For example, the Korean PSI search unit 153 performs a search in the Korean resource database 154 for a search word translated into Korean and an authority heading in Korean from the Korean authority heading extraction unit 151. In addition, the English PSI search unit 164 performs a search in the English resource database 164 for a search word translated in English and an authorized authority heading in English from the English authority head extracting unit 161, and the Japanese PSI search. The unit 174 performs a search in the Japanese resource database 174 for the Japanese translation of the search word and the Japanese authority head from the Japanese authority head extracting unit 171.

Each of the plurality of resource databases 154, 164, and 174 stores and manages resources constructed ontologies based on a topic map structure according to a corresponding language. Multiple resource databases 154, 164, 174 store data in a library's book retrieval system, a web site information retrieval system, a small network connected to a local area network (LAN), or a personal computer (PC). Various resources can be provided in conjunction with one storage device. In particular, such a storage device may store and manage a large amount of distributed information about a document file, an image file, a video file, database information, or text data, and the information is more accurate and rich to the user through semantic search. For the purpose of providing, it may be information constructed by ontology connected to a map (topic map) of semantic relations centered on a topic.

Topic map elements for a topic map ontology include a topic type, an association type, an association role type, an occurrence type, an occurrence type, and a published subject. Identifier: public subject indicator) may include input elements and connection lines (connection lines).

In the topic map ontology, the topic type may be represented in a circle, and the name of the topic type may be entered in the circle. Topics are abstract concepts that categorize topics such as 'composer', 'opera', and 'musician', as shown in FIGS. It can be any word or phrase on any subject that can be a subject.

The association type is represented in the form of a rhombus or diamond, and the name of the association type may be entered therein. As with Figs. 4 and 5, the association type specifies the correlation between the topic types by causing the topic types to be connected by connection lines at the corners. Association relationships represent various semantic relationships as well as vertical relationships between topic types.

When connecting the association type to the topic type, the triangular association roll type is set.In order to show the role of the association type, the name of the roll type in the triangle and the label content in the parentheses are inserted. Help to convey meaning As shown in Figs. 4 and 5, the arrow of the connection line indicates the direction of meaning transfer between topics. For example, in FIGS. 4 and 5, there is an association of 'composed by' between 'composer' and 'opera', and 'composer' is composed by 'work', that is, 'opera' ', And' opera 'refers to the relationship including the' composer '(composed).

The occurrence type is represented by a hexagon, and the name of the occurrence type can be entered therein. An occurrence is an internal or external resource associated with the topic, and the occurrence resource has a data type such as 'string', 'number', 'date', 'Uniform Resource Identifier' (URI), or 'Uniform Resource Locator' (URL). It may include. For example, the occurrence data type can be a URI that points to the location of a resource outside the topic map, and can take various forms, such as document files, image files, video files, information in a database, or text data that a topic can reference. have. At this time, all occurrences for a topic share all attributes represented by the topic. As shown in FIG. 6, occurrence types included in a topic type may be connected by a dotted line among the connection lines.

Each of the topic type, association type, association role type, and occurrence type on the topic map ontology can be described with a published subject identifier (PSI). The PSI input element is represented by a solid line, and the solid end point For example, public topic indicators can be defined as 'URI' or 'URL' to share the topic. Search engines can interpret the unique value of a topic from PSI information.

As such, the ontology may be built on the basis of RDF, RDFS, OWL, etc. for various types of information accessible to the user, such as information managed on a personal computer or information shared on a network such as the Internet. Their search or navigation is free, easy and efficient.

The resource databases 154, 164, and 174 manage resources constructed using the topic map ontology. Each of the plurality of PSI search units 153, 163, and 173 is based on a search word translated into a corresponding language. For a plurality of authority headings in the language, each search is started to perform a semantic search.

The PSI search units 153, 163, and 173 may search based on the order of character arrangement of the language and perform one search according to the search rank. Among these, PSI having at least one frequency of each search word or authorized heading may be extracted.

For example, the Korean PSI search unit 153 may be a sequence of characters arranged in Korean when a search word translated in a corresponding language and a plurality of authority headings in a corresponding language are “Da Vinci” or “Leonardo” as shown in FIG. 3. Search for 'DaVinci' first, according to (Ganada. Or order), then extract the PSI for the resource with the frequency of the word, and then search for 'Leonardo' in the same way. PSI can be extracted for a resource having a frequency of the corresponding word. If there are no related words in the resources, it is possible to search for a resource with similar words or limit the number of search results if there are too many resources with related words. Similarly, the English PSI searcher 163 and the Japanese PSI searcher 173 may perform a search on the translated search word and the corresponding authority headings in a similar manner as above.

The search result providing unit 180 collects the search results for each language extracted by the plurality of PSI search units 153, 163, and 173 as described above (S260) to provide summary information according to a predetermined display order. It may be (S270). As shown in FIG. 7, the search result providing unit 180 may provide a title and a corresponding summary of search results for a corresponding language in a predetermined language order, for example, Korean, English, Japanese, and Chinese order according to a character arrangement order. Can be. For each language, the search results for the search word or the corresponding translation can be provided above and the search results for the authority heading below.

In addition, according to the present invention, it is possible to provide a search word input window to input a logical combination (including AND or OR) of a plurality of search words on a screen as shown in FIG. 8 or 9 through a predetermined user interface. 8 or 9 illustrates an example of inputting two search terms, more logical combinations of search terms are possible, and the user attempts to search for more accurate information search results by inputting a plurality of search terms in the search term input window. can do.

When a plurality of search terms of the logical combination method are input, the multilingual translator 110 may perform a translation on all of the input plurality of search terms to generate a plurality of search words translated in each language for each search term. Accordingly, the plurality of PSI search units 153, 163, and 173 may be authorized for each of the plurality of search words translated in a corresponding language and the plurality of search terms from the plurality of authority head extracting units 151, 161, and 171, respectively. The heading may be searched in a database based on a corresponding language among the plurality of resource databases 154, 164, and 174.

In this case, each of the plurality of PSI search units 153, 163, and 173 may include a plurality of search words translated into a corresponding language for a plurality of search terms in a logical combination method and a plurality of search words in a corresponding language corresponding to each of the translated plurality of search words. As described above, the authority headings are searched based on the order of character arrangement of the language, and searched one by one according to the corresponding search rank. However, for each resource extracted by performing the respective search on the corresponding translated multiple search terms and corresponding authority headings, while satisfying the corresponding logical combination (such as AND or OR), the corresponding translated multiple search terms and corresponding authority headings PSIs having frequencies of at least one can be extracted.

For example, when searching with 'secret' and 'da vinci code' as OR, as shown in FIG. 8 or 9, the Korean PSI search unit 153 searches for the term 'secret' and its authority heading ' For each of "A" (assuming the authorized authority title is A) and "Davinci code" and its authorized "'B" (assuming that the authorized authority title is B), perform one search according to its search rank. Can be. In this case, the Korean PSI search unit 153 may extract PSIs of resources having a frequency of at least one or more, such that 'Da Vinci code', 'Secret', authority head 'A', authority head 'B', and the like.

As illustrated in FIG. 8 or 9, the user interface may provide a corresponding weight setting window next to a plurality of search terms inputted in a logical combination as described above. The weight may be set by dragging and dropping a bar-shaped setting mark of a display window with a mouse as shown in FIG. 8, or by inputting a corresponding percentage or rank with a number as shown in FIG. 9.

When the weight is set in this way, the search result providing unit 180 may control the user interface to display the search results on the screen according to the weight order set in the weight setting window. For example, in the example of FIG. 8, when the weight of 'secret' is higher than the weight of 'davinci code', a search result for 'secret' may be first displayed on the screen as shown in FIG. 10. Alternatively, in the example of FIG. 9, when the weight of 'secret' is lower than the weight of 'davinci code', a search result for 'davinci code' may be first displayed on the screen as shown in FIG. 11. Although not shown in FIG. 10 or 11, when there is a foreign resource for 'secret' or 'davinci code', search results may be provided for each language, and if there are authorized authority headings for each language Corresponding search results may also be provided.

In addition, the user interface may provide a viewing step setting window on the search result screen as shown in FIGS. 12, 13, and 14, and the search result providing unit may set the depth of the search result in the viewing step set in the viewing step setting window. The user interface may be controlled to display separately. For example, in providing the information of the materials stored in the library in the library user system, as shown in FIG. 12, in the first step, the summary of the title, author name, publisher, and publication year of the book is included in the search result. Information can be provided. Alternatively, in the second stage, additional summary information such as the index information of the library and the current state (rental, processing, etc.), which includes the contents of the first stage, may be provided. In Example 3, additional summary information, including the content of Example 2, and more detailed information, such as the title, author name, publisher, and year of publication, may be provided.

As described so far, in the ontology-based semantic retrieval method according to the present invention, first, the authority headings are databased for each word type for a plurality of words in each country language, and the topic map structure according to each country language is used. On the basis of this, resources are built ontologies and databased. The database constructed as such is the basis of the semantic search according to the present invention.

When a search word is input, a plurality of search words translated in each country language of a plurality of countries may be generated for the search word, and each country is recognized and received by recognizing a search word in each country language among the plurality of translated search words. It is determined whether there is an authority heading corresponding to the search word in each language by extracting the authority heading with reference to the authority heading databased for the language search word. Once the authority heading is extracted, a search is performed on the database resources for the searched word translated into the corresponding language and the authority heading in each language, and the search results for each language are collected and summarized according to a predetermined display order. Can be provided.

As described above, optimal embodiments have been disclosed in the drawings and the specification. Although specific terms have been used herein, they are used only for the purpose of describing the present invention and are not intended to limit the scope of the invention as defined in the claims or the claims. Therefore, those skilled in the art will understand that various modifications and equivalent other embodiments are possible from this. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

1 is a view for explaining an ontology-based semantic search system according to an embodiment of the present invention.

FIG. 2 is a conceptual diagram for describing an operation of the ontology-based semantic retrieval system of FIG. 1.

3 is an example of authority headings.

4 is an example showing a part of a topic map structure such as a topic type, an association type, an association roll type, and a PSI element.

5 is another example showing a part of a topic map structure such as a topic type, an association type, an association roll type, and a PSI element.

6 is another example illustrating a part of a topic map structure such as a topic type, an occurrence type, a PSI element, and the like.

7 is an example of a search result according to an embodiment of the present invention.

8 is an example of a search word input window for weighting a search condition.

9 is another example of a search word input window for weighting a search condition.

10 is an example of a search result when a weight is given to the search word 'secret'.

11 illustrates an example of a search result when a weight is given to the search term 'DaVinci code'.

12 is an example of example 1 of a search result.

FIG. 13 is an example of a step 2 example of search results. FIG.

14 is an example of three viewing examples of search results.

Claims (9)

A multilingual translator configured to generate a plurality of search terms translated in each country language of the plurality of countries with respect to the input search word; A plurality of language recognition units for recognizing and receiving a search word in a corresponding language among the translated plurality of search words; A plurality of authority heading databases for storing and managing authority headings for each word type for a plurality of words in a corresponding language; A plurality of resource databases each storing and managing resources constructed ontologies based on a topic map structure according to a corresponding language; For each of the search terms in the corresponding language from the plurality of language recognition units, it is determined whether there is an authority head corresponding to the search term in the corresponding language by referring to a database based on the corresponding language among the plurality of authority title databases. A plurality of authority head extraction units for extracting authority heads;  A plurality of search units which perform a search in a database based on a corresponding language among the plurality of resource databases with respect to a search word translated into a corresponding language and authority heads in a corresponding language from the plurality of authority head extracting units; And A search result providing unit for collecting the search results for each language extracted by the plurality of search units and providing summary information according to a predetermined display order; Ontology-based semantic retrieval system comprising a. The method of claim 1, wherein each of the plurality of search units, For a search word translated in a corresponding language and a plurality of authority headings in the corresponding language, a search rank is determined based on the character arrangement order of the corresponding language, and the search is performed one by one according to the search rank. An ontology-based semantic search system for extracting PSIs having at least one frequency of each search term or authority head, from among resources including each search term or authority head, when performing a search. The method of claim 1, wherein the plurality of authority heading databases, respectively, An ontology-based semantic retrieval system comprising a database of authorized headings according to word forms including names of persons, names, other proper nouns, common nouns, adjectives / adverbs, or verbs. The method of claim 1, It further includes a user interface for providing a search term input window to enter a logical combination (including AND or OR) of multiple search terms on the screen, The multilingual translator translates all of the plurality of search terms to generate a plurality of search words translated in each language for each search term. The plurality of search units may be a database based on a corresponding language among the plurality of resource databases for each authority head corresponding to each of the plurality of search terms translated in a corresponding language and the plurality of search terms from the plurality of authority head extractors. An ontology-based semantic search system, characterized in that to perform a search. The method of claim 4, wherein each of the plurality of search units, A plurality of search words translated in the corresponding language and a plurality of authority headings in the corresponding language corresponding to each of the plurality of search terms are searched one by one according to the search rank by determining a search rank based on a character arrangement order of the corresponding language. And then Extracting PSIs having at least one frequency of each of the search terms or authorized headings while satisfying the logical combinations among resources including each search term or authorized headings for a plurality of search terms inputted through the logical combinations when performing a search An ontology-based semantic search system. The method of claim 4, wherein The user interface provides a weight setting window for the plurality of search terms, And the search result providing unit controls the user interface to display the search results on the screen according to the weight order set in the weight setting window. The method of claim 6, The user interface provides a viewing step setting window on a search result screen. And the search result providing unit controls the user interface to classify and display the depth of the search result according to the viewing step set in the viewing step setting window. The method of claim 1, wherein the plurality of resource databases, A large volume of distributed information about document files, image files, video files, database information, or text data on small networked or personal PCs connected to a library, book retrieval systems in libraries, Web site information retrieval systems, and personal computers. An ontology-based semantic retrieval system characterized by providing resources in association with a storage means storing data constructed by ontology connected to a map of semantic relations. Database the authority headings for each word type for a plurality of words in each language; Constructing and databasening resources as ontology based on a topic map structure according to each language; Generating a plurality of search words translated in each country language of the plurality of countries with respect to the input search word; Recognizing and receiving a search word in each language among the translated plurality of search words; Determining whether there is an authority heading corresponding to the search term in each country language by extracting the authority heading with reference to the database of authority headings for the search term in each country language; Performing a search on the database resources for a search term translated in a corresponding language and authority headings in each language; And Collecting search results for each language extracted by performing the search and providing summary information according to a predetermined display order; Ontology-based semantic search method comprising a.

Images