Abstract:
PURPOSE: An ontology based instance identification system and a method thereof are provided to improve the accessibility of information through deduction between meanings by generating an ontology instance which a URI(Uniform Resource ID) is given when converts data of an RDBMS(Relational Database Management System) into the instance of an OWL(Web Ontology Language) ontology. CONSTITUTION: An ontology register registers an ontology consisting of a class, an attribute, the relationship information between the classes in an ontology database(810) of a database unit. If a target instance is inputted, a class-instance linker(400) selects and connects the class which the instance is allocated. A reference/additional information obtainer(600) obtains the information required for the identification of the target instance. A controller(500) identifies the target instance from the ontology database by using the information.
Abstract:
PURPOSE: An ontology based sameAs management system and a method thereof are provided to share the set up including generation or deletion of a sameAs relation about ontology instances with different ID by synchronizing sameAs management servers with a equal level on a real time basis and synchronizing the high and lower level sameAs management server. CONSTITUTION: Storage(111) stores the set up information including generation or deletion of a sameAs relation and the relationship information for checking the sameAs relation. A relation turnaround unit(113) checks the sameAs relation of an instance corresponding to a query language of a user and returns the relation of a sameAs management server. A query generating unit(115) generates an SPARQL(Simple Protocol and RDF(Resource Description Framework) Query Language) query which the instance within a list is reflected. A result generating unit(117) generates a result corresponding to the SPARQL query.
Abstract:
PURPOSE: A substitute reference solution system and method thereof are provided to improve realizing ratio of core object recognition by distinguishing core object referring to pronoun. CONSTITUTION: An anaphoric pronoun recognizing module(200) recognizes a pronoun having preceding word through the alternative pronoun, the neighboring lexical information and part of speech column information in document. A controller(100) receives the pronoun information. An antecedent extracting module(300) receives pronoun information to find a preceding word which a pronoun infers to. A pronoun recognition module uses a selection of vocabulary unit(210) and unit of a speech selecting unit(220) and recognizes a pronoun in the document.
Abstract:
PURPOSE: A semantic syntactic tree kernel based processing system and method thereof are provided to automatically extract an associative relationship by calculating similarity of two sentences by a semantic syntactic tree. CONSTITUTION: A language analysis unit(110) analyzes CRF(conditional random field) information. A integration quality processing unit(120) receives an analysis result. A similarity extracting unit(130) calculates similarity by calculating a semantic parse tree kernel by the relation of syntax. A similarity extracting unit(130) calculates similarity and semantic syntactic tree kernel.
Abstract:
본 발명은 정보검색명령 및 자료관리명령 등의 정보명령을 입력하는 사용자 단말기; 상기 사용자 단말기로부터 상기 정보명령을 입력받아 상기 정보명령에 따른 작업 수행을 분배하여 지시하는 작업지시 신호를 출력하는 작업관리 수단; 상기 작업지시 신호를 입력받아 정보검색을 수행하는 복수개의 정보검색 수단; 및 상기 정보검색 수단에 연결되며, 다수의 데이터를 저장하고 있는 복수개의 데이터 베이스를 포함하며, 상기 작업관리 수단은 상기 복수개의 정보검색 수단 중 작업을 하지 않는 정보검색 수단을 선별하여 작업지시를 하도록 하는 작업 지시 신호를 출력하며, 상기 복수개의 정보검색 수단은 상기 복수개의 데이터 베이스에 각각 연결되고, 연결된 데이터 베이스에 대해 정보검색을 각각 수행하는 복수개의 쓰레드를 포함하는 것을 특징으로 하는 정보검색관리시스템을 제공함으로써 대용량의 작은 데이터 갱신을 안정적으로 수행할 수 있는 정보검색이 가능하도록 한다.
Abstract:
PURPOSE: A full text XML treatise construction system and method are provided to automatically convert various treatise files into a full text XML treatise suitable for a style of a treatise defined in data type definition(DTD) of a PubMed central(PMC). CONSTITUTION: A screen input unit(100) inputs text information for a treatise for research performance in a PDF document by a paragraph unit through a text input part on a web input page. An item setting unit(110) controls the text input part by mapping the text information with an item defined in DTD of a PMC. When the item is selected, an automatic change unit(120) changes a stored type of text data in the text input part corresponding to the selected item. [Reference numerals] (100) Screen input unit; (110) Item setting unit; (120) Data attribute automatic change unit; (130) XML thesis generating unit; (AA) Database storage; (BB) Outputting a web screen; (CC) Setting a text field item; (DD) Storing data; (EE) Generating Full-Text XML; (FF) Mapping PMC DTD