公开(公告)号：KR101881731B1

公开(公告)日：2018-07-25

申请号：KR1020180023585

申请日：2018-02-27

Applicant: 한국과학기술정보연구원

Inventor： 전홍우 ,  김희철 ,  김선호 ,  김정준 ,  고병열 ,  권오진 ,  문영호

IPC: G16H50/30 ,  G16H50/70 ,  A61B5/00 ,  G06N99/00

CPC classification number: G16H50/30 ,  A61B5/4088 ,  A61B5/7275 ,  G06N99/005 ,  G16H50/70

Abstract: 본발명은치매예측을위하여기계학습장치로입력될사용자의연도별의료데이터를처리하며, 최적의자질들로조합된데이터세트가구성된다. 최적의자질들에는적어도사용자의병력(Disease History)에관한정보가포함되고, 사용자의최근 7년이하의연도별의료정보가포함된다. 사용자의연도별의료데이터에서실험을통해파악된최적의자질들을구성함으로써정확한치매의예측과진단을할 수있다. 구체적인실험결과는무조건오랜기간동안의의료정보를관찰하기보다는최근 7년이하의병력을관찰한예측결과가가장뛰어난것을보여주므로, 치매예측에적정한기준을제시한다.

公开(公告)号：KR101374195B1

公开(公告)日：2014-03-13

申请号：KR1020110123596

申请日：2011-11-24

Applicant: 한국과학기술정보연구원

Inventor： 전홍우 ,  최성필 ,  최윤수 ,  정창후 ,  성원경 ,  송사광 ,  이원구 ,  정도헌

IPC: G06F17/21 ,  G06F17/30

Abstract: 다차원 지식을 자동으로 추출하고 활용 및 공유할 수 있는 대규모 학술적 과학 지식 메모리 기반의 심층 지식 제공을 위한 방법 및 장치가 개시된다.
학술적 과학 지식 메모리 기반의 심층 지식 제공 장치는 문서에 대한 관계형 지식(Relational Knowledge), 구조적 지식(Structural Knowledge), 절차적 지식(Procedural Knowledge)을 저장하는 지식 메모리; 및
질의어를 입력하고, 상기 지식 메모리로부터 입력된 질의어를 포함하는 트리플 및 트리플에 관련된 문서들을 검색하여 제시하는 심층 지식 제공부;를 포함한다.

公开(公告)号：KR101233423B1

公开(公告)日：2013-02-15

申请号：KR1020120052735

申请日：2012-05-17

Applicant: 한국과학기술정보연구원

Inventor： 최성필 ,  전홍우 ,  정창후 ,  최윤수 ,  정한민

IPC: G06F17/21 ,  G06F17/27 ,  G06F17/30

CPC classification number: G06F17/27

Abstract: PURPOSE: A device for extracting a descriptive concept expression about terms from literature and a method thereof are provided to actually enable an conceptual approach to the literature by performing a sentence search based on a PAT(Predicate Argument Tuble) for extracting a sentence corresponding to an inputted term. CONSTITUTION: A conversion processing unit(100) designates a correlation about words included in each paragraph of a sentence based on a grammatical relation between the words composing the sentence to extract a PAT corresponding to the sentence. The conversion processing unit reforms a structure of the PAT for grouping the PAT. An information storage unit(200) stores the reformed PAT so that the PAT corresponds to the sentence. An information search unit(300) extracts the stored sentence to include the PAT extracted from a sentence corresponding to a query. [Reference numerals] (100) Conversion processing unit; (200) Information storage unit; (300) Information search unit; (AA) Literature; (BB) User query language

Abstract translation: 目的：提供一种用于从文献中提取关于术语的描述性概念表达的装置及其方法，以便通过执行基于PAT（谓词参数Tulum）的句子搜索来实际地实现对文献的概念性方法，以提取与 输入期限。 构成：转换处理单元（100）基于构成句子的词之间的语法关系来指定包含在句子的每个段落中的词的相关性，以提取对应于该句子的PAT。 转换处理单元改变用于分组PAT的PAT的结构。 信息存储单元（200）存储重整的PAT，使得PAT对应于该句子。 信息搜索单元（300）提取存储的句子以包括从与查询相对应的句子中提取的PAT。 （附图标记）（100）转换处理单元; （200）信息存储单元; （300）信息搜索单位; （AA）文学; （BB）用户查询语言

公开(公告)号：KR1020120068076A

公开(公告)日：2012-06-27

申请号：KR1020100104765

申请日：2010-10-26

Applicant: 한국과학기술정보연구원

Inventor： 최성필 ,  최윤수 ,  정창후 ,  성원경 ,  전홍우

IPC: G06F17/20 ,  G06F17/30

CPC classification number: G06F17/277 ,  G06F17/2735 ,  G06F17/2785

Abstract: PURPOSE: A method and an apparatus for automatically detecting the technological concept of terminology from technical references are provided to interlink new technology and existing technology. CONSTITUTION: A linguistic meaning symbol extracting part(102) respectively extracts the meaning expression of technical terms and the meaning expression of amplification description. An implication relation verifying part(104) verifies the implication relation of the meaning expressions of the technical terms and the amplification description. A new linguistic expression detecting part(106) extracts existing technical concepts to the technical terms from technology references. The new linguistic expression detecting part detects new technological concepts or new technical terms based on the technical concepts.

Abstract translation: 目的：提供一种从技术参考中自动检测术语技术概念的方法和装置，用于互联新技术和现有技术。 构成：语言意义符号提取部分（102）分别提取技术术语的含义表达和放大描述的含义表达。 含义关系验证部分（104）验证技术术语的含义表达式与放大描述的含义关系。 一种新的语言表达检测部分（106）从技术参考文献中提取技术术语的现有技术概念。 新的语言表达检测部分基于技术概念检测新技术概念或新技术术语。

公开(公告)号：KR101693014B1

公开(公告)日：2017-01-05

申请号：KR1020140150893

申请日：2014-11-03

Applicant: 한국과학기술정보연구원

Inventor： 여운동 ,  이재민 ,  전홍우 ,  김선호 ,  송인석 ,  고병열

IPC: G06Q10/06 ,  G06Q50/18 ,  G06Q40/04

Abstract: 본발명은빅데이터를활용하여기업의제품개발전략을효과적으로벤치마킹하기위한제품그물망구축장치, 그방법및 기업의제품개발전략을벤치마킹하는제품그물망구축프로그램을저장하는저장매체를제공하기위함이다. 이를위해본 발명은, 적어도하나이상의기업에대하여, 연도별로등록된적어도하나이상의특허및 연도별로등록된적어도하나이상의특허에대응하는적어도하나이상의제품에대한정보를입력받는단계(S90), 특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품의제 1 비중정보를획득하는단계(S91), 특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품중 특정제품군에해당하는제품의제 2 비중정보를획득하는단계(S92), 제 1 비중정보와제 2 비중정보를이용하여특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품과특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품중 특정제품군에해당하는제품간의파생관계를나타내는파생정보를계산하는단계(S93), 특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품을보유한기업이상기특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품중 특정제품군에해당하는제품을보유할수 있는확률을지시하는전방다각화확률값을구하는단계(S94), 특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품중 특정제품군에해당하는제품을보유한기업이특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품을보유하였을확률을지시하는후방다각화확률값을구하는단계(S95)를포함한다.

Abstract translation: 本发明涉及一种用于通过使用大数据对公司的产品开发策略进行基准化的产品链制作装置，其方法和用于存储用于基准公司的产品开发策略的产品链制作程序的存储介质。 产品链制作装置包括：输入模块; 第一获取模块，被配置为获得与特定年份中注册的至少一个专利对应的至少一个产品的第一重要性信息; 第二获取模块，被配置为获得与特定年份中注册的至少一个专利对应的一个或多个产品的产品的第二重要性信息，其中产品对应于特定产品组; 和产品多样化模块。 因此，该设备使用户能够有效地找到产品开发策略。

公开(公告)号：KR1020140016784A

公开(公告)日：2014-02-10

申请号：KR1020120085253

申请日：2012-08-03

Applicant: 한국과학기술정보연구원

Inventor： 전홍우 ,  최성필 ,  정창후 ,  황미녕 ,  정성재 ,  정한민

IPC: G06F19/28

CPC classification number: G06F19/28 ,  G06F17/30716 ,  G06F17/30731

Abstract: The present invention relates to a system and method for expanding a pathway. The present invention includes a pathway database which stores pathways which are previously formed and the system for expanding the pathway, which generates a relation event by analyzing a document if the document is newly inputted and updates the pathway according to a comparison result after the generated relation event is compared with the pathways stored in the pathway database. [Reference numerals] (100) Pathway forming system; (110) Pre-information DB; (120) Relationship information DB; (140) Object recognition unit; (150) Relationship recognition unit; (160) Relations event recognition unit; (170) Pathway update unit; (180) Visualization unit; (190) Validation unit; (200) Pathway DB

Abstract translation: 本发明涉及用于扩展路径的系统和方法。 本发明包括存储先前形成的路径的路径数据库和用于扩展路径的系统，其通过在文档被新输入时分析文档来生成关系事件，并且在生成关系之后根据比较结果来更新路径 事件与存储在途径数据库中的路径进行比较。 （附图标记）（100）路径形成系统; （110）预信息DB; （120）关系信息DB; （140）对象识别单元; （150）关系识别单位; （160）关系事件识别单位; （170）路径更新单元; （180）可视化单位; （190）验证单位; （200）路径DB

公开(公告)号：KR101243063B1

公开(公告)日：2013-03-13

申请号：KR1020120085254

申请日：2012-08-03

Applicant: 한국과학기술정보연구원

Inventor： 전홍우 ,  최성필 ,  정창후 ,  황미녕 ,  정성재 ,  정한민

IPC: G06F19/28

CPC classification number: G06F17/30696 ,  G06F17/30864 ,  G06F19/12 ,  G06F19/24

Abstract: PURPOSE: A pathway construction system and a method thereof are provided to recognize words expressing protein, diseases, enzymes, medicines, compounds, and symptoms from a biofield-related document, thereby automatically constructing a pathway. CONSTITUTION: An entity recognition unit(140) recognizes words by performing morpheme analysis, syntax analysis, and semantic analysis for an input document. A relation recognition unit(150) recognizes a relation between entities by extracting context between the entities and making the context as a pattern defined in a normalized dictionary database. A related event generation unit(160) generates a related event by performing a web search for the entities. The related event generation unit generates the related event by collecting a document including the entities and space information in cells of the entities. A pathway generation unit(170) generates a pathway by displaying the entities in a space of a cell based on the related event. [Reference numerals] (100) Pathway construction system; (110) Dictionary information DB; (120) Relation information DB; (130) Pathway DB; (140) Entity recognition unit; (150) Relation recognition unit; (160) Related event generation unit; (170) Pathway generation unit; (180) Visualization unit; (190) Verification unit

Abstract translation: 目的：提供一种途径构建系统及其方法，从生物场相关文件中识别表达蛋白质，疾病，酶，药物，化合物和症状的词，从而自动构建途径。 构成：实体识别单元（140）通过对输入文档执行语素分析，语法分析和语义分析来识别单词。 关系识别单元（150）通过提取实体之间的上下文并将上下文作为在归一化字典数据库中定义的模式来识别实体之间的关系。 相关事件生成单元（160）通过对这些实体执行网络搜索来生成相关事件。 相关事件生成单元通过在实体的单元中收集包括实体和空间信息的文档来生成相关事件。 路径生成单元（170）通过基于相关事件在小区的空间中显示实体来生成路径。 （附图标记）（100）通道施工系统; （110）词典信息DB; （120）关系信息DB; （130）通路DB; （140）实体识别单元; （150）关系识别单元; （160）相关事件发生单元; （170）途径生成单位; （180）可视化单位; （190）验证单元

公开(公告)号：KR101240330B1

公开(公告)日：2013-03-11

申请号：KR1020120077375

申请日：2012-07-16

Applicant: 한국과학기술정보연구원

Inventor： 정창후 ,  최성필 ,  전홍우 ,  김태홍 ,  엄정호 ,  정성재 ,  정한민

IPC: G06F17/21 ,  G06F17/30

CPC classification number: G06F17/3071 ,  G06F17/2795

Abstract: PURPOSE: A multi-dimensional document classification system and a method thereof are provided to improve document classification efficiency by applying a document classification algorithm to a group generated by grouping classification systems. CONSTITUTION: A classification system generation unit(210) generates a part classification system based on stored learning target documents. A quality extraction unit(220) extends qualities by extracting the qualities from the learning target document of the part classification system. A classifier generation unit(230) generates a learning model for the part classification system by using the extracted or extended qualities. When a new document is inputted, a document classification unit(250) classifies the document based on a learning model including the qualities by extracting the qualities from the inputted document. [Reference numerals] (200) Classifier; (210) Classification system generation unit; (220) Quality extraction unit; (230) Classifier generation unit; (240) Learning model DB; (250) Document classification unit

Abstract translation: 目的：提供一种多维文档分类系统及其方法，通过将文档分类算法应用于通过分组分类系统生成的组来提高文档分类效率。 构成：分类系统生成单元（210）基于存储的学习目标文档生成部分分类系统。 质量提取单元（220）通过从部件分类系统的学习目标文档中提取质量来扩展质量。 分类器生成单元（230）通过使用提取的或扩展的质量来生成用于分类系统的学习模型。 当输入新文档时，文档分类单元（250）通过从输入的文档中提取质量，基于包括质量的学习模型对文档进行分类。 （附图标记）（200）分类器 （210）分类系统生成单元; （220）质量提取单位; （230）分类器生成单元; （240）学习模型DB; （250）文件分类单位

公开(公告)号：KR1020120047377A

公开(公告)日：2012-05-14

申请号：KR1020100108666

申请日：2010-11-03

Applicant: 한국과학기술정보연구원

Inventor： 최윤수 ,  최성필 ,  전홍우 ,  정창후 ,  성원경

IPC: G06F17/27 ,  G06F17/25

CPC classification number: G06F17/27 ,  G06F17/25

Abstract: PURPOSE: A feature generating method for recognizing an entity of machine learning is provided to generate a word unit feature suitable of a named entity recognition system. CONSTITUTION: A word in named entity dictionary is discriminated in a paragraph. The kind of word in named entity dictionary is discriminated. A word unit having the kind of named entity is given. The quality of the word by word is generated. The feature of the word unit, which is generated from the named entity dictionary is integrated.

Abstract translation: 目的：提供一种用于识别机器学习实体的特征生成方法，用于生成适用于命名实体识别系统的单词特征。 构成：命名实体字典中的一个词在一段中被区分开来。 命名实体字典中的这种词被歧视。 给出了具有命名实体种类的单元。 生成单词的质量。 从命名实体字典生成的单词单元的特征被集成。

公开(公告)号：KR1020160051264A

公开(公告)日：2016-05-11

申请号：KR1020140150893

申请日：2014-11-03

Applicant: 한국과학기술정보연구원

Inventor： 여운동 ,  이재민 ,  전홍우 ,  김선호 ,  송인석 ,  고병열

IPC: G06Q10/06 ,  G06Q50/18 ,  G06Q40/04

CPC classification number: G06Q10/067 ,  G06Q40/04 ,  G06Q50/18

Abstract: 본발명은빅데이터를활용하여기업의제품개발전략을효과적으로벤치마킹하기위한제품그물망구축장치, 그방법및 기업의제품개발전략을벤치마킹하는제품그물망구축프로그램을저장하는저장매체를제공하기위함이다. 이를위해본 발명은, 적어도하나이상의기업에대하여, 연도별로등록된적어도하나이상의특허및 연도별로등록된적어도하나이상의특허에대응하는적어도하나이상의제품에대한정보를입력받는단계(S90), 특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품의제 1 비중정보를획득하는단계(S91), 특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품중 특정제품군에해당하는제품의제 2 비중정보를획득하는단계(S92), 제 1 비중정보와제 2 비중정보를이용하여특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품과특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품중 특정제품군에해당하는제품간의파생관계를나타내는파생정보를계산하는단계(S93), 특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품을보유한기업이상기특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품중 특정제품군에해당하는제품을보유할수 있는확률을지시하는전방다각화확률값을구하는단계(S94), 특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품중 특정제품군에해당하는제품을보유한기업이특정연도에등록된적어도하나이상의특허각각에대응하는적어도하나이상의제품을보유하였을확률을지시하는후방다각화확률값을구하는단계(S95)를포함한다.

Abstract translation: 本发明涉及一种用于通过使用大数据对公司的产品开发策略进行基准化的产品链制作装置，其方法和用于存储用于基准公司的产品开发策略的产品链制作程序的存储介质。 产品链制作装置包括：输入模块; 第一获取模块，被配置为获得与特定年份中注册的至少一个专利对应的至少一个产品的第一重要性信息; 第二获取模块，被配置为获得与特定年份中注册的至少一个专利对应的一个或多个产品的产品的第二重要性信息，其中产品对应于特定产品组; 和产品多样化模块。 因此，该设备使用户能够有效地找到产品开发策略。