WO2017150820A1 - Système d'expansion de graphe conceptuel fondé sur une base de connaissances - Google Patents

Système d'expansion de graphe conceptuel fondé sur une base de connaissances Download PDF

Info

Publication number
WO2017150820A1
WO2017150820A1 PCT/KR2017/001592 KR2017001592W WO2017150820A1 WO 2017150820 A1 WO2017150820 A1 WO 2017150820A1 KR 2017001592 W KR2017001592 W KR 2017001592W WO 2017150820 A1 WO2017150820 A1 WO 2017150820A1
Authority
WO
WIPO (PCT)
Prior art keywords
concept
graph
triple
conceptual
concept graph
Prior art date
Application number
PCT/KR2017/001592
Other languages
English (en)
Korean (ko)
Inventor
최성필
Original Assignee
경기대학교 산학협력단
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 경기대학교 산학협력단 filed Critical 경기대학교 산학협력단
Publication of WO2017150820A1 publication Critical patent/WO2017150820A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor

Definitions

  • the present invention relates to a question and answer technique, and more particularly to a paraphrase generation technique for a query.
  • An object of the present invention is to provide a technical solution that can help the comprehensiveness and scalability of the question and answer.
  • the concept graph extension system is extended by using the reference information stored in the relationship triple knowledge base based on the relation triple knowledge base storing the reference information for the concept graph expansion and the concept graph converted from a query composed of natural language sentences. It may include a conceptual graph extension module that generates a conceptual graph of the form.
  • the concept graph may include concept nodes representing at least some objects constituting the query and relationship nodes representing a relationship between the concept nodes.
  • the concept graph extension module can expand a concept graph around a concept node and iteratively expand an extended concept graph around a concept node.
  • the concept graph extension system may visualize the concept graph and provide it through a user interface, and may further include a concept graph extension visualization module that works with the concept graph extension module.
  • the reference information stored in the relationship triple knowledge base may include information about a concept entity and a relationship triple representing a relationship between the concept entities.
  • the relation triple knowledge base may have a schema structure for storing relational knowledge data.
  • the database schema structure of the relationship triple knowledge base consists of a first specification containing concept labels and definition statements and links, a second specification containing attributes and relationship labels and definition statements linking concepts, and a set of relationship triples for objects.
  • the first information and the second information may be stored, the third information including a relationship between the attributes, and the fourth information designating a range of values to be stored in the attributes.
  • the concept graph extension module selects one or more concept objects to extend from the query triplet set in the query concept graph, extracts a triple set from the definition statements of the selected concept object, and uses the selected definition triples from the extracted definition statement triple sets. You can extend the concept graph.
  • the concept graph extension module may select a concept object of high importance from the concept objects belonging to the question triple set.
  • the concept graph extension module compares the definition triple and the question triple set to remove duplicate definition triples, selects the definition triple that matches the subject triple or subject or object, and selects the selected definition triple. You can select a triple definition statement that matches the subject and the object or object. You can remove all remaining triples.
  • query paraphrase and extension that can help the comprehensiveness and extensibility of the query response based on the concept graph are possible.
  • FIG. 1 is a block diagram for a conceptual graph-based query translation and expansion according to an embodiment.
  • FIG. 2 is an exemplary diagram illustrating a conceptual graph for a specific question.
  • 3 is an exemplary diagram related to query conceptual graph extension.
  • FIG. 4 is a conceptual diagram of a query conceptual graph extension system according to an exemplary embodiment.
  • FIG. 5 shows an example of a database schema structure for storing wiki data.
  • FIG. 6 is a block diagram illustrating a knowledgebase based conceptual graph extension system according to an exemplary embodiment.
  • FIG. 7 illustrates an example of a conceptual graph extended input / output test bed screen.
  • FIG. 10 is a block diagram illustrating a query concept graph extension system to which a query concept graph extension visualization module is added, according to an exemplary embodiment.
  • FIG. 11 shows an example of a conceptual graph extension visualization module configuration.
  • 13 is an exemplary view showing the initial expansion of the concept graph.
  • FIG. 14 is an exemplary diagram illustrating an additional conceptual node extension for “Big Bang”.
  • FIG. 1 is a block diagram for paraphrase and extension of a conceptual graph based query according to an embodiment
  • FIG. 2 is an exemplary diagram illustrating a conceptual graph for a specific question
  • FIG. 3 is an exemplary diagram related to conceptual graph expansion.
  • the concept graph generation module 100 receives a query sentences composed of natural language and generates a concept graph (CG). Since this concept graph is a concept graph generated from a query statement, it can be called query concept graph (Query CG). According to an embodiment, the concept graph generation module 100 may generate a concept graph by using a triple structure of a query sentence having subject, predicate or relation, and object information. have.
  • the concept graph may be composed of concept nodes and relationship nodes representing relationships between the concept nodes.
  • the subject and object are designated as concept nodes, and the relation is designated as relationship nodes.
  • the concept graph generation module 100 may grasp the triple structure of the query sentence, and generate the concept graph with the subject, the relationship, and the object.
  • the concept graph generation module 100 said, “Carbon emission trading system was introduced to regulate the emission of greenhouse gases causing global warming, and Korea plans to implement it next year. What is the international agreement that laid the foundation for the carbon emission trading system ”can be generated in the form of a query, as shown in FIG. This is called the original concept graph.
  • the concept graph extension module 200 receives the original concept graph from the concept graph generation module 100 and expands it.
  • the concept graph expansion module 200 converts the input original concept graph into a semantic triple set and then performs a concept graph expansion algorithm to be described later.
  • Conceptual graph extension algorithm can be performed based on the knowledge base.
  • the concept graph extension module 200 may derive the sentences written on the left side of FIG. 3 based on the semantic triple set and knowledge base converted from the original concept graph, and from this, the triple as shown in the table on the right side of FIG. A set can be created, and an extended conceptual graph can be generated based on the generated triple set.
  • FIG. 4 is a conceptual diagram of a conceptual graph extension system according to an exemplary embodiment.
  • Conceptual graphs contain a variety of information, including entity names such as person names, place names, organization names, etc., as well as their relationships, attribute names, and attribute values.
  • subject and object are entities (conceptual entities), and relationships can be called predicates.
  • the disclosed system has a flexible structure that can be extended to all these elements. All the elements that constitute the nodes and connections of the generated conceptual graph can be assumed to be semantically distinct and unambiguous. However, the disclosed system can be used to provide context and ambiguity resolution information so that it can be appropriately extended depending on the context. It is configured to be processed by additional input.
  • various reference information may be applied according to a situation. For example, it is possible to implement a conceptual graph extension system based on relational knowledge data such as wiki data.
  • the conceptual graph extension algorithm performed in the disclosed system is as follows.
  • the original conceptual graph entered is converted into a question triple set (meaning triple set).
  • the conceptual graph expansion module 200 selects terms to be extended in the set of questions triples.
  • the concept graph expansion module 200 selects concepts having high importance from concept entities of the concept graph.
  • the concept graph extension module 200 selects the terms listed in the wiki data among the concept entities of the concept graph as having high importance.
  • the concept graph extension module 200 may select subjects and objects selected and provided from the concept graph generation module 100 as concepts having high importance. 2
  • the concept graph expansion module 200 extracts a triple set from the definition statement of the picked-up object, and compares the extracted definition statement triple set and the question triple set.
  • the concept graph expansion module 200 returns to the process 1 according to the predetermined number of expansion iterations. In other words, if the number of extended iterations is one, it goes back to 1 process only, and if the number of extended iterations is two times, it goes back to 1 process twice. The number of extended repetitions may be specified by the user.
  • FIG. 5 shows an example of a database schema structure for storing wiki data.
  • the database structure utilized by the Wikidata-based concept graph extension system is illustrated in FIG. 5.
  • the first description includes a concept label, a definition syntax, and a link indicating an object, that is, a single page of wiki data
  • the second description includes an attribute and a relationship label connecting the concepts.
  • definition syntax For reference, the meaning of the object herein does not mean the object described above, but is a term used in Wikidata, and means an object to be described by a specific Wikipedia page. For example, if you search for a Wikipedia page with the term "information search", the page corresponding to this object is found, and the searched page describes the object.
  • the first information (Information 1) and the second information (Information 2) store a relationship triple set for an object
  • the third information (Information 3) includes a relationship between attributes. It is.
  • the fourth information (Information 4) specifies a range of values to be stored in the attribute.
  • the disclosed modular database schema is designed to accommodate not only wiki data but also various types of semantic triple databases. In particular, it is easy to refer to additional information and surrounding information for a specific concept to facilitate identification of different concepts having the same label (name).
  • FIG. 6 is a block diagram illustrating a knowledgebase based conceptual graph extension system according to an exemplary embodiment.
  • FIG. 6 shows a detailed configuration diagram of a system that sequentially expands concept nodes, attributes, etc. belonging to a concept graph in the relationship triple knowledge base 300 based on a given concept graph.
  • the concept graph extension system may include a concept graph extension module 200 and a relationship triple knowledge base 300.
  • the relationship triple knowledge base 300 stores reference information for expanding a concept graph, and the reference information includes information about a concept entity and a relationship triple representing a relationship between the concept entities.
  • the concept entity may mean a subject and an object.
  • the database schema structure of the relationship triple knowledge base 300 is as shown in FIG.
  • the conceptual graph expansion module 200 includes a user connection module 210 and a relation triple expansion module 220.
  • the user access module 210 may include a user interface 211 and a Java package interface, through which a user may call a desired function using the concept graph extension module 200 and receive a result.
  • the relation triple extension module 220 queries and queries the relation triple knowledge base 300 and obtains the knowledge base triple (relationship triple) as a result.
  • the relationship triple extension module 220 may identify an object ID of a label, identify a label of an object ID, identify a definition statement of an object, identify a label of an attribute ID, identify an ID of an attribute label, and attributes. Identify definition statements, identify object-object relationship structures, identify object-value relationship structures, filter to refined relationships, and identify wiki links for objects.
  • the triple-based conceptual graph expansion module 200 includes various types of functions. First of all, there is a module for identifying identifiers in the knowledge base 300 or vice versa for object names (labels or conceptual names). There are also functions. Detailed API specification is shown in Table 1.
  • an identifier for the node's name (concept name, entity name, object name) is needed.
  • all nodes of the conceptual graph are semantically discriminated.
  • the semantic ambiguity can be removed by mapping to a single meaning in the process of generating a concept graph and thus can have only one identifier.
  • the concept of “universe” can have a total of 23 different meanings in Wikidata as below, but it is designated as “label # 1” in the concept graph and entered as the input of this system.
  • the output of the disclosed system may be stored and provided in JSON form.
  • an extension I / O test bed in a simple client form for directly verifying the operation of the system is configured as shown in FIG. 7 so that an extension result for a specific query concept can be directly checked.
  • 1 denotes a concept name input portion
  • 2 denotes an attribute input portion
  • 3 denotes an extension range setting portion.
  • 4 is the result output part. If you enter a name or identifier and attribute corresponding to a specific concept, and an extended scope, the expanded concept is finally displayed.
  • a definition statement (English and Korean) briefly defining a concept corresponding to an input identifier is output first, and various extension concepts having a specific relationship with the input concept are output below.
  • each relationship name is also managed as an identifier, so these identifiers can be output together.
  • an additional definition statement is provided for the extended concepts with various relationships. If the extension range is increased from 1 to 2, the secondary concepts connected with the first expanded concepts may be additionally output.
  • Certain concepts in Wikidata are linked to other concepts as well as to additional information such as category information, topic names, thesaurus identifiers, image files, and input dates.
  • Figure 8 shows the expansion result for the "universe” discussed earlier.
  • “q1” denotes a specific identifier of “universe” for convenience, and English and Korean definition sentences thereof are output at the top.
  • the various kinds of relation triples that contain "universe” as the subject are printed. These triples are the result of searching in real time directly within the database and exist in relations such as “P793” and “P31”. For example, "universe"-"significant event”-"Big Bang” refer to “Big Bang” in a "space” and “significant event” relationship.
  • the disclosed system may output triples based on all relations depending on options, or output only a triple set corresponding to the relation specified in the concept graph.
  • the system includes multiple extension functions that mean more than 2nd extensions, it is more effective to avoid 2nd or more extensions as much as possible because there is a risk of deriving unnecessary information in terms of actual query conceptual graphs. have.
  • JSON files output by APIs and contents of a specific JSON file.
  • "Q1_ID_O_Triple_i.json” stores all primary triple sets that have an input concept as a subject.
  • the actual JSON file contains only the identifiers corresponding to the subject, relation, and object. Refer to the detailed information (label, English / Korean definition, wiki page number, etc.) for the identifier. To do this, use the associated API.
  • a conceptual graph extension module 400 is added to the query concept graph extension system as shown in FIG. Can be linked with (200).
  • the system can be called a query concept graph extension visualization system by distinguishing it from the query concept graph extension system, or it can be called a query concept graph extension system.
  • the concept graph extension visualization module 400 may visualize a concept graph expressed as a relationship between nodes and provide them to a user through the user interface 211.
  • the conceptual graph extension visualization module 400 may be developed based on the graph visualization module.
  • An example of a conceptual graph extension visualization module 400 is shown in FIG. 11. 11 illustrates an AJAX-based web visualization module and a client-based visualization module developed by using the concept graph extension module 200 described above.
  • WebVOWL 0.5.x http://vowl.visualdataweb.org/webvowl.html
  • OWL Web Ontology Language
  • FIG. 12 shows an example of the entire screen configuration of WebVOWL.
  • the left part of FIG. 12 is an area where a conceptual graph that can be continuously expanded is visualized and interacts with a user.
  • On the right side there is an information providing area in which the title, description, statistics, and detailed information on the selected node are output. Through this, the user can check specific information about a specific concept node and relationship node.
  • “critical event” as a relation node is selected, and detailed information about “critical event” is output in the information providing area.
  • the node may be selected by the user.
  • there is a utility area at the bottom of the screen to perform operations such as screen manipulation, external export, gravity adjustment, filtering, and reassembly.
  • Existing WebVOWL takes a single static and static OWL file as input and visualizes it, whereas the disclosed system has been reconfigured to output a constantly expanding dynamic conceptual graph.
  • FIG. 13 is an exemplary diagram illustrating an initial expansion of a conceptual graph
  • FIG. 14 is an exemplary diagram illustrating an additional conceptual node expansion for a “big bang”
  • FIG. 15 is an example of a detailed extended conceptual graph for “Gorge Le Metre”.
  • These figures show the concept graph continuously expanding, starting with a specific subject (“universe”), showing the characteristics of the disclosed system. As shown in FIG. 13, if the original original concept graph includes a concept node called "universe,” it can be extended around a specific relationship ("critical event") to obtain new additional concept nodes. . Examples include “expansion theory”, “big bang nuclear synthesis”, and “big bang”.
  • the related concept nodes include "general relativity” and “opposite concept” formed in a “total-part” relationship as shown in FIG. Relational "normal cosmology” nodes, and “discoverer” relations, can be derived.
  • the results of performing the concept node extension for “Gorgeous Le Metre” are shown in FIG. 15. As shown in FIG. 15, when there are a plurality of object nodes for a specific relationship, virtual nodes are configured and connected.
  • Performance evaluation criteria are as follows. To select the incorrect triple entry, first of all, the relationship between the two conceptual words is awkward triples (“Smart City”-“kind”-“movie”), and triples whose overall meaning may be unclear or different if they are expanded as part of the concept graph. "Social Network”-"Higher Classification”-”Social Construct”), triples ("barcode”-"type”-”machine-readable”) whose object meaning is indeterminate, and conceptual graph-based question and answer systems Considered all inappropriate triples ("Massachusetts Institute of Technology"-"website account”-”researchgate”) to be incorrect triples. Of the total 142 triplets, 25 false triplets were present and the accuracy was calculated to be 78.63%.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

La présente invention concerne un système d'expansion de graphe conceptuel. Le système peut comprendre : une base de connaissances de triple relation destinée à stoker les informations de référence pour l'expansion d'un graphe conceptuel; et un module d'expansion de graphe conceptuel destiné à générer un graphe conceptuel dans une forme expansée à l'aide des informations de référence stockées dans la base de connaissances de triple relation, sur la base d'un graphe conceptuel converti à partir d'une interrogation exprimée en tant que phrase de langage naturel.
PCT/KR2017/001592 2016-02-29 2017-02-14 Système d'expansion de graphe conceptuel fondé sur une base de connaissances WO2017150820A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
KR10-2016-0024383 2016-02-29
KR1020160024383A KR101835345B1 (ko) 2016-02-29 2016-02-29 지식베이스 기반의 개념그래프 확장 시스템

Publications (1)

Publication Number Publication Date
WO2017150820A1 true WO2017150820A1 (fr) 2017-09-08

Family

ID=59743062

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2017/001592 WO2017150820A1 (fr) 2016-02-29 2017-02-14 Système d'expansion de graphe conceptuel fondé sur une base de connaissances

Country Status (2)

Country Link
KR (1) KR101835345B1 (fr)
WO (1) WO2017150820A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110457435A (zh) * 2019-07-26 2019-11-15 南京邮电大学 一种专利新颖性分析系统及其分析方法
WO2024007119A1 (fr) * 2022-07-04 2024-01-11 华为技术有限公司 Procédé d'apprentissage de modèle de traitement de texte, ainsi que procédé et dispositif de traitement de texte
US11960513B2 (en) 2019-09-18 2024-04-16 Saltlux Inc. User-customized question-answering system based on knowledge graph

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101934583B1 (ko) * 2018-01-17 2019-01-02 주식회사 머니브레인 대화형 ai 에이전트 시스템을 위한 지식베이스의 시각화 방법 및 컴퓨터 판독가능 기록 매체
KR102090237B1 (ko) * 2018-07-31 2020-03-17 주식회사 포티투마루 시맨틱 트리플 기반의 지식 확장 시스템, 방법 및 컴퓨터 프로그램
KR102261199B1 (ko) * 2018-09-19 2021-06-07 주식회사 포티투마루 인공 지능 질의 응답 시스템, 방법 및 컴퓨터 프로그램
KR102128549B1 (ko) 2018-09-19 2020-07-08 주식회사 포티투마루 인공 지능 질의 응답 시스템, 방법 및 컴퓨터 프로그램
US11562133B2 (en) 2018-12-04 2023-01-24 Foundation Of Soongsil Univ-Industry Cooperation System and method for detecting incorrect triple
KR102212310B1 (ko) * 2018-12-04 2021-02-05 숭실대학교산학협력단 오류 트리플 검출 시스템 및 방법
KR102206742B1 (ko) 2018-12-06 2021-01-25 한국과학기술원 자연언어 텍스트의 어휘 지식 그래프 표현 방법 및 장치
CN110532328B (zh) * 2019-08-26 2023-04-07 哈尔滨工程大学 一种文本概念图构造方法
KR102439165B1 (ko) * 2019-11-26 2022-09-01 한국과학기술원 상식 지식과 언어학적 패턴을 이용한 서술문 신뢰도 평가 시스템 및 그 방법
KR102398832B1 (ko) * 2020-04-23 2022-05-17 주식회사 케이티 지식 그래프에 기초하여 응답을 도출하는 장치, 방법 및 컴퓨터 프로그램
KR102655154B1 (ko) * 2021-09-30 2024-04-05 주식회사 와이즈넛 인공신경망 모델을 이용한 컨셉넷 자동 확장 방법
WO2023249558A1 (fr) * 2022-06-22 2023-12-28 Gp Network Asia Pte. Ltd. Procédé et système pour exécuter de manière adaptative une pluralité de tâches

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100820746B1 (ko) * 2007-01-22 2008-04-11 조선대학교산학협력단 온톨로지를 이용한 정보의 브라우징 시스템 및 방법
KR20130125999A (ko) * 2012-05-10 2013-11-20 경북대학교 산학협력단 키워드 추출 방법 및 시스템, 그리고 대화 보조 장치
KR20130126114A (ko) * 2012-05-10 2013-11-20 최진근 단어 연관성을 이용하는 정보 수집 시스템 및 정보 수집 방법
KR20140051606A (ko) * 2012-10-23 2014-05-02 에스케이텔레콤 주식회사 Rdf 기반의 문장 온톨로지를 이용한 일관성 평가 방법 및 장치
JP2015505082A (ja) * 2011-12-12 2015-02-16 インターナショナル・ビジネス・マシーンズ・コーポレーションＩｎｔｅｒｎａｔｉｏｎａｌ Ｂｕｓｉｎｅｓｓ Ｍａｃｈｉｎｅｓ Ｃｏｒｐｏｒａｔｉｏｎ 情報ドメインに対する自然言語処理モデルの生成

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100820746B1 (ko) * 2007-01-22 2008-04-11 조선대학교산학협력단 온톨로지를 이용한 정보의 브라우징 시스템 및 방법
JP2015505082A (ja) * 2011-12-12 2015-02-16 インターナショナル・ビジネス・マシーンズ・コーポレーションＩｎｔｅｒｎａｔｉｏｎａｌ Ｂｕｓｉｎｅｓｓ Ｍａｃｈｉｎｅｓ Ｃｏｒｐｏｒａｔｉｏｎ 情報ドメインに対する自然言語処理モデルの生成
KR20130125999A (ko) * 2012-05-10 2013-11-20 경북대학교 산학협력단 키워드 추출 방법 및 시스템, 그리고 대화 보조 장치
KR20130126114A (ko) * 2012-05-10 2013-11-20 최진근 단어 연관성을 이용하는 정보 수집 시스템 및 정보 수집 방법
KR20140051606A (ko) * 2012-10-23 2014-05-02 에스케이텔레콤 주식회사 Rdf 기반의 문장 온톨로지를 이용한 일관성 평가 방법 및 장치

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110457435A (zh) * 2019-07-26 2019-11-15 南京邮电大学 一种专利新颖性分析系统及其分析方法
US11960513B2 (en) 2019-09-18 2024-04-16 Saltlux Inc. User-customized question-answering system based on knowledge graph
WO2024007119A1 (fr) * 2022-07-04 2024-01-11 华为技术有限公司 Procédé d'apprentissage de modèle de traitement de texte, ainsi que procédé et dispositif de traitement de texte

Also Published As

Publication number Publication date
KR20170101609A (ko) 2017-09-06
KR101835345B1 (ko) 2018-03-07

Similar Documents

Publication Publication Date Title
WO2017150820A1 (fr) Système d'expansion de graphe conceptuel fondé sur une base de connaissances
US6618727B1 (en) System and method for performing similarity searching
CN108446367A (zh) 一种基于知识图谱的包装行业数据搜索方法及设备
Lopez et al. Powermap: Mapping the real semantic web on the fly
Rinaldi et al. A matching framework for multimedia data integration using semantics and ontologies
WO2021049706A1 (fr) Système et procédé de réponse aux questions d'ensemble
Prudhomme et al. Interpretation and automatic integration of geospatial data into the Semantic Web: Towards a process of automatic geospatial data interpretation, classification and integration using semantic technologies
WO2011122730A1 (fr) Système et procédé destinés à un service de recherche connexe basé sur un réseau rdf
Zhao et al. Mid-ontology learning from linked data
Jetinai et al. Ontology mapping and rule-based inference for learning resource integration
Zhang et al. Construction of fuzzy ontologies from fuzzy XML models
WO2012046904A1 (fr) Procédé et dispositif pour fournir des informations de recherche à partir de ressources multiples
WO2018147543A1 (fr) Système de questions-réponses basé sur un graphe de concept et procédé de recherche de contexte l'utilisant
Aaberge et al. Ontology and Ontology Construction: Background and Practices.
Suryanarayana et al. Stepping towards a semantic web search engine for accurate outcomes in favor of user queries: Using RDF and ontology technologies
Fahad et al. Disjoint-knowledge analysis and preservation in ontology merging process
Preethi et al. Case and relation (CARE) based page rank algorithm for semantic web search engines
Myint et al. Triple patterns extraction for accessing data on ontology
Zhang et al. Semantic web and geospatial unique features based geospatial data integration
Kubis An access layer to PolNet–polish WordNet
Tang et al. Ontology-based semantic retrieval for education management systems
CN111506640A (zh) 映射方法及装置
Maeda et al. CoMeMo: Constructing and sharing everyday memory
Lauw et al. TUBE (Text-cUBE) for discovering documentary evidence of associations among entities
Stanley et al. A conceptual-kdd approach and its application to cultural heritage

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17760224

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 17760224

Country of ref document: EP

Kind code of ref document: A1