-
公开(公告)号：WO2012050252A1
公开(公告)日：2012-04-19
申请号：PCT/KR2010/007243
申请日：2010-10-21
CPC classification number: G06F17/30705 , G06K9/00288
Abstract: 본 발명은 분류기의 동적 결합에 의한 대용량 분류기 자동 생성 시스템 및 방법에 관한 것으로서, 문서의 특성에 따라 다수의 문서가 분류되어 저장되는 데이터베이스, 각 데이터베이스의 학습 대상 문서로부터 자질을 추출하고, 상기 추출된 자질과 범주의 유사도를 구하여 자질별 가중치를 연산한 후, 각 자질에 대해 가중치를 포함하는 자질 특성 매트릭스 및 용어 벡터를 생성하여 신규로 수집되는 분류대상 문서의 범주를 결정하는 개별 분류기를 생성하는 분류기 생성 장치, 복수개의 결합 대상 자질 특성 매트릭스에 출현한 자질들의 코드정보를 리스트화하고, 각 자질에 대한 범주 특성값을 합산하여 통합 매트릭스를 생성한 후 상기 통합 매트릭스를 이용하여 결합 대상 데이터베이스의 모든 자질 정보를 학습한 통합 분류기를 생성하는 분류기 동적 결합 장치를 포함한다. 따라서, 본 발명에 따르면, 대용량의 문서를 학습시킬 수 있는 대용량 분류기 생성을 위해 복수의 결합대상 데이터베이스를 동적으로 결합하여 어떤 데이터베이스에 대해서도 범용적으로 적용할 수 있다.
Abstract translation: 本发明涉及一种使用分类器的动态组合自动生成质量分类器的系统和方法，包括：存储根据文档特征分类的多个文档的数据库; 从数据库的每个学习对象文档中提取质量的分类器生成装置计算提取的质量和类别之间的相似度，从而计算每个质量的加权，然后生成包括每个质量的加权的项向量和质量特征矩阵 以生成确定新分类对象文档的类别的单个分类器; 以及用于列出关于要组合的多个质量特征矩阵的质量的代码信息的分类器的动态组合的装置，将用于质量的类别特征值相加，并且生成积分矩阵以生成积分分类器，其具有 了解要组合的数据库的所有质量信息。 因此，根据本发明，动态地组合多个数据库以产生允许大量文档被学习的大容量分类器，并且其可以通常应用于任何数据库。
-
公开(公告)号：WO2017142109A1
公开(公告)日：2017-08-24
申请号：PCT/KR2016/001645
申请日：2016-02-18
Applicant: 한국과학기술정보연구원
CPC classification number: G06F17/30371 , G06F17/2765 , G06F17/30241 , G06F17/30516 , G06F17/30551 , G06F17/30569 , G06F17/30684 , G06F17/30699 , G06N99/005
Abstract: 본 발명은 비정형 텍스트 추출 성능 향상을 위한 시스템 및 방법에 관한 것으로서, 수집된 비정형 텍스트를 언어 분석하여, 이벤트 키워드, 이벤트가 발생한 시간정보 또는 공간정보를 추출하고, 상기 이벤트 키워드에 시간정보 또는 공간정보를 매핑시켜 추출 지식 후보들을 생성하는 비정형 데이터 처리부, 시공간 연계 정형 데이터를 이용하여 상기 비정형 데이터 처리부에서 생성된 추출 지식 후보들의 유효성을 판단하는 필터부를 포함한다.
Abstract translation:
本发明涉及一种用于改进的提取自由文本性能的系统和方法，对所收集的自由文本，事件的关键字，以及该事件的语言分析是提取时间信息或空间信息发生时， 映射在事件关键字的时间信息或空间信息，以包括所述过滤器的部件，以确定使用非结构化数据处理，空时联的结构化数据，以生成知识提取候选通过非结构化数据处理单元生成的提取的知识的候选人的有效性。
-
公开(公告)号：KR101739018B1
公开(公告)日：2017-06-08
申请号：KR1020170008699
申请日：2017-01-18
Applicant: 한국과학기술정보연구원
CPC classification number: G06Q30/0269 , G06F17/27 , G06Q30/0202 , G06Q30/0244 , G06Q30/0261
Abstract: 본발명은고객선정장치로기구축된데이터베이스로부터고객정보및 가맹점정보를수신하여프로파일정보를생성하는프로파일정보생성모듈; 마케팅타겟가맹점정보를생성하는타겟가맹점선정모듈; 상기생성된프로파일정보를기반으로상기마케팅타겟가맹점정보에대응하는잠재적구매고객정보를생성하는잠재고객정보생성모듈; 상기생성된잠재적구매고객정보를기반으로매출기여도정보를생성하는타겟고객선정모듈; 을포함한다.
Abstract translation: 本发明组，以产生由该客户的信息和从数据库中建立的简档信息生成模块给顾客选择装置的商户信息接收到的简档信息; 目标商家选择模块，用于生成营销目标商家信息; 潜在顾客信息生成模块，用于基于所生成的简档信息生成对应于所述营销目标商家信息的潜在购买顾客信息; 目标顾客选择模块，用于基于所生成的潜在购买顾客信息生成销售贡献信息; 它包括。
-
公开(公告)号：KR101195492B1
公开(公告)日：2012-10-29
申请号：KR1020110133705
申请日：2011-12-13
Applicant: 한국과학기술정보연구원
CPC classification number: G06F17/30696 , G06F17/30657 , G06F17/30867
Abstract: PURPOSE: A query result expecting method and a system thereof are provided to supply information for a query word which users input as an information search keyword and a subject range for an information search result. CONSTITUTION: Query result expecting information about a query word is requested(S404). Subject field distribution about the query word transmits to a user terminal as the query result expecting information(S408). The query result expecting information is outputted in a predetermined area of a search screen(S412). A search requesting signal including the query word transmits to a search service providing server(S428). The search service providing server searches information about the query word(S440). [Reference numerals] (AA) User terminal; (BB) Start; (CC) Search service providing device; (DD,FF,HH) Yes; (EE,GG,II) No; (JJ) End; (S402) Query word input; (S404) Requesting query result predicting information; (S406) Obtaining related word information or subject field distribution for a query word; (S408) Generating query result predicting information including the related word information or the subject field distribution; (S410) Query result predicting information; (S412) Query result predicting information output; (S414) Query equation change?; (S416) Requesting the subject field distribution; (S418) Obtaining the subject field distribution for the related word included in a query equation; (S420) Renewing the subject field distribution by combining the subject field distribution for the related word and the subject field distribution for the query word; (S422) Renewed subject field distribution; (S424) Renewed subject field distribution output; (S426) Subject field change?; (S428) Search request signal; (S430) Filtering only information for subject fields selected by considering the query word; (S432) Search result; (S434) Search result output; (S436) Search command input?; (S438) Search request signal; (S440) Information search for the query word; (S442) Search result; (S444) Search result output
Abstract translation: 目的：提供一种查询结果预期方法及其系统，用于为用户作为信息搜索关键字输入的查询词和信息搜索结果的对象范围提供信息。 构成：请求查询结果，期望查询词的信息（S404）。 关于查询字的主题分布将作为查询结果期望信息发送到用户终端（S408）。 在搜索屏幕的预定区域中输出期望信息的查询结果（S412）。 包括查询字的搜索请求信号发送到搜索服务提供服务器（S428）。 搜索服务提供服务器搜索关于查询词的信息（S440）。 （附图标记）（AA）用户终端; （BB）开始; （CC）搜索服务提供设备; （DD，FF，HH）是; （EE，GG，II）否; （JJ）结束; （S402）查询字输入; （S404）请求查询结果预测信息; （S406）获取查询词的相关词信息或主题字段分布; （S408）生成包括相关字信息或被摄体分布的查询结果预测信息; （S410）查询结果预测信息; （S412）查询结果预测信息输出; （S414）查询方程改变？ （S416）请求主体分配; （S418）获取包含在查询方程中的相关词的主题分布; （S420）通过组合相关词的主题分布和查询词的主题分布来更新主题分配; （S422）重新科目现场分配; （S424）更新科目现场分配产出; （S426）主题字段更改？ （S428）搜索请求信号; （S430）仅筛选通过考虑查询词所选择的主题字段的信息; （S432）搜索结果; （S434）搜索结果输出; （S436）搜索命令输入 （S438）搜索请求信号; （S440）查询词的信息搜索; （S442）搜索结果; （S444）搜索结果输出
-
公开(公告)号：KR1020100054588A
公开(公告)日：2010-05-25
申请号：KR1020080113565
申请日：2008-11-14
Applicant: 한국과학기술정보연구원
IPC: G06F17/30
CPC classification number: G06F17/2785
Abstract: PURPOSE: A method for extracting a semantic associative relation of context based on syntax analysis of whole document, a device thereof and a recording medium storing a program source are provided to extract the semantic associative relation of context based on syntax analysis of whole document, thereby enabling to connect to UPK(Undiscovered Public Knowledge). CONSTITUTION: A TRS(Tech Retrieval System)(110) manages whole data which recognition of the technical terms are completed and provides searching service in different modules. A TAS(Tech Acquisition System)(150) searches the technical terms based on a dictionary from data of the TRS and extracts new terms. A TLA(Tech Language Analyzer)(180) does chunking and tagging the source document to conveniently extract the relation between technical terms recognition and relation. A TAMA(Tech Association Mining Appliance)(170) analyzes the document by construction through the TLA and extracts a semantic associative relation of context and performs conceptualizing based on word net. A SATT(Semi-Automatic Tech-Tracking Engine)(160) has semantic associative relation and provides many services.
Abstract translation: 目的：提供基于整个文档的语法分析，其设备和存储程序源的记录介质的上下文语义关联关系的方法，以基于整个文档的语法分析提取上下文的语义关联关系，由此 能够连接到UPK（未被发现的公共知识）。 规定：TRS（技术检索系统）（110）管理完整的技术术语识别数据，并提供不同模块的搜索服务。 TAS（技术采购系统）（150）根据TRS数据的字典搜索技术术语，并提取新术语。 TLA（技术语言分析器）（180）对源文档进行分块和标记，方便地提取技术术语识别与关系之间的关系。 TAMA（技术协会采矿设备）（170）通过TLA构建分析文档，并提取语境的语义关联关系，并基于字网进行概念化。 SATT（半自动技术跟踪引擎）（160）具有语义关联关系，提供许多服务。
-
Patent Agency Ranking
公开(公告)号：KR1020090114779A
公开(公告)日：2009-11-04
申请号：KR1020080040596
申请日：2008-04-30
Applicant: 한국과학기술정보연구원
IPC: G06F17/30
CPC classification number: G06F17/30539 , G06F17/2785 , G06F2216/03
Abstract: PURPOSE: An extracting device for a semantic technological term of a high-capacity database is provided to rapidly perform technical examination, examination, and decision by searching/tracing the associative relations of the technological information, time series analysis, and classification in a real time. CONSTITUTION: An ARM(Acquisition Rule Manager) unit extracts a query language. The extracted query language is provided as a TRS(Tech Retrieval System) unit(S110). The TRS unit provides the document group of the designated technology and posting information to an analysis unit(S120). The analysis unit registers the new and seed technological term to the knowledge database(S130,S140). The document group and the information of the knowledge database is provided to a tracking unit and the related technology is traced. The traced information is outputted with a document.
Abstract translation: 目的：提供大容量数据库语义技术术语提取装置，通过实时搜索/跟踪技术信息，时间序列分析和分类的关联关系，快速进行技术检查，检查和决策 。 构成：ARM（采集规则管理器）单元提取查询语言。 提取的查询语言作为TRS（技术检索系统）单元提供（S110）。 TRS单元向分析单元提供指定技术的文档组和发布信息（S120）。 分析单元将新的种子技术术语登记到知识数据库（S130，S140）。 将文档组和知识数据库的信息提供给跟踪单元，并追踪相关技术。 跟踪的信息与文档一起输出。
-
公开(公告)号：KR1020090065626A
公开(公告)日：2009-06-23
申请号：KR1020070133017
申请日：2007-12-18
Applicant: 한국과학기술정보연구원
Abstract: A method and a system for providing the information of a study which is suitable to a user concerning the relative field of the user based on a level are provided to reduce the difficulty of metabolism researchers generated in the study start in a special field. A communication unit(210) performs communication with a client. If interesting field selection information is received through the communications unit, a detailed field search unit(230) extracts the detailed field information about the field of interest. The detailed field searcher produces a user profile information input picture including the detailed field list among the extracted detailed field information. If the user profile information is received from client, a study order information generating unit(240) performs the filtering operation based on inside of the user profile information the special knowledge extent. The research order information generator produces the research order information about the field of interest.
Abstract translation: 提供了一种方法和系统，用于提供适合于用户基于用户级别的用户的相关领域的研究的信息的方法和系统，以减少在特定领域中在研究开始时产生的代谢研究人员的难度。 通信单元（210）执行与客户端的通信。 如果通过通信单元接收到感兴趣的字段选择信息，则详细字段搜索单元（230）提取关于感兴趣字段的详细字段信息。 详细的字段搜索器产生包括提取的详细字段信息中的详细字段列表的用户简档信息输入画面。 如果从客户端接收到用户简档信息，则学习订单信息生成单元（240）基于用户简档信息的特定知识范围进行过滤操作。 研究订单信息生成器产生关于感兴趣领域的研究订单信息。
-
公开(公告)号：KR101644429B1
公开(公告)日：2016-08-10
申请号：KR1020160018386
申请日：2016-02-17
Applicant: 한국과학기술정보연구원
CPC classification number: G06F17/30371 , G06F17/2765 , G06F17/30241 , G06F17/30516 , G06F17/30551 , G06F17/30569 , G06F17/30684 , G06F17/30699 , G06N99/005
Abstract: 본발명은비정형텍스트추출성능향상을위한시스템및 방법에관한것으로서, 수집된비정형텍스트를언어분석하여, 이벤트키워드, 이벤트가발생한시간정보또는공간정보를추출하고, 상기이벤트키워드에시간정보또는공간정보를매핑시켜추출지식후보들을생성하는비정형데이터처리부, 시공간연계정형데이터를이용하여상기비정형데이터처리부에서생성된추출지식후보들의유효성을판단하는필터부를포함한다.
Abstract translation: 本发明涉及一种非结构化文本提取性能改进系统及其方法。 该系统包括：非结构化数据处理单元，被配置为语言分析收集的非结构化文本以提取事件关键字;以及时间信息或空间信息，其中事件被生成并被配置为将事件关键字中的时间信息或空间信息映射 产生提取知识候选人; 以及滤波器单元，被配置为通过使用时空相关结构化数据来确定在非结构化数据处理单元中生成的提取知识候选的有效性。
-
公开(公告)号：KR101259418B1
公开(公告)日：2013-04-30
申请号：KR1020120052262
申请日：2012-05-17
Applicant: 한국과학기술정보연구원
CPC classification number: G06Q50/10 , G06F17/30 , G06F17/30734
Abstract: PURPOSE: An adaptive user guide service method and a server are provided to find potential concerns through collecting and analyzing real time user feedback, thereby supplying a desired result to a user. CONSTITUTION: A user feedback information collection unit(240) collects user feedback information from user terminals through a communication unit(230). An entity extraction unit(250) extracts entities from the collected user feedback information. A transfer candidate service selection unit(260) selects a transfer candidate service with a precondition corresponding to information related to an extracted entity among transfer candidate services of a first service through a search a service guide information database(220). A service providing unit(270) transfers the first service to the selected transfer candidate service. [Reference numerals] (200) Service providing server; (210) Service information DB; (220) Service guide information DB; (230) Communication unit; (240) User feedback information collection unit; (250) Entity extraction unit; (260) Transfer candidate service selection unit; (270) Service providing unit; (280) Ontology DB;
Abstract translation: 目的：提供自适应用户指南服务方法和服务器，通过收集和分析实时用户反馈来发现潜在的问题，从而为用户提供所需的结果。 构成：用户反馈信息收集单元（240）通过通信单元（230）从用户终端收集用户反馈信息。 实体提取单元（250）从所收集的用户反馈信息中提取实体。 转移候选服务选择单元（260）通过搜索服务指南信息数据库（220），选择具有与第一服务的转移候选服务中提取的实体相关的信息的前提条件的转移候选服务。 服务提供单元（270）将第一服务传送到所选择的转移候选服务。 （附图标记）（200）服务提供服务器; （210）服务信息DB; （220）服务指南信息DB; （230）通讯单元; （240）用户反馈信息收集单元; （250）实体提取单元; （260）转移候选服务选择单元; （270）服务提供单位; （280）本体数据库;
-
公开(公告)号：KR101035038B1
公开(公告)日：2011-05-19
申请号：KR1020100099164
申请日：2010-10-12
Applicant: 한국과학기술정보연구원
CPC classification number: G06F17/30705 , G06K9/00288
Abstract: PURPOSE: A system and method for automatically generating high capacity of classifier by a dynamic combination of a classifier are provided to freely generates a unified classifier which learnd all quality information of combined target databases by dynamically combining a plurality of combining target databases. CONSTITUTION: Databases(100a~100c) classifies and store documents. Classification generating devices(200a~200n) generates an individual classifier. The individual classifier calculates a quality weighted vale by obtaining a similarity of range of quality extracted from a learning target document of the database. The individual classifier determines a range of classified target documents by generating a feature matrix and term vector including the weight value. A classifier dynamic combining device(300) generates an integrated classifier which learned all quality information nof a combination target database.
Abstract translation: 目的：提供一种通过分类器的动态组合自动生成分类器容量的系统和方法，以自由生成统一的分类器，通过动态组合多个组合目标数据库，学习组合目标数据库的所有质量信息。 规定：数据库（100a〜100c）对文件进行分类和存储。 分类生成装置（200a〜200n）生成单个分类器。 单个分类器通过获得从数据库的学习目标文档中提取的质量范围的相似度来计算质量加权值。 单个分类器通过生成包括权重值的特征矩阵和项向量来确定分类的目标文档的范围。 分类器动态组合装置（300）生成综合分类器，该分类器学习组合目标数据库的所有质量信息。
-
-
-
-
-
-
-
-
-
Search Results
28
records
(took 0.024 seconds)
