WO2023169072A1 - 知识图谱中实体的配置方法、分析方法及装置 - Google Patents

知识图谱中实体的配置方法、分析方法及装置 Download PDF

Info

Publication number
WO2023169072A1
WO2023169072A1 PCT/CN2023/070481 CN2023070481W WO2023169072A1 WO 2023169072 A1 WO2023169072 A1 WO 2023169072A1 CN 2023070481 W CN2023070481 W CN 2023070481W WO 2023169072 A1 WO2023169072 A1 WO 2023169072A1
Authority
WO
WIPO (PCT)
Prior art keywords
entity
expression
logical
knowledge graph
entities
Prior art date
Legal status (The legal status 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 status listed.)
Ceased
Application number
PCT/CN2023/070481
Other languages
English (en)
French (fr)
Inventor
何雨潇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Alipay Hangzhou Information Technology Co Ltd
Original Assignee
Alipay Hangzhou Information Technology Co Ltd
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 Alipay Hangzhou Information Technology Co Ltd filed Critical Alipay Hangzhou Information Technology Co Ltd
Publication of WO2023169072A1 publication Critical patent/WO2023169072A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/35—Clustering; Classification
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/02—Knowledge representation; Symbolic representation
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/04—Inference or reasoning models

Definitions

  • One or more embodiments of this specification relate to the computer field, and in particular, to a configuration method, analysis method and device for entities in a knowledge graph.
  • Ontology knowledge can be represented by SPO (Subject-Predicate-Object) triples, which are used to describe the relationship between two different entities (concepts or objects). Relationship. Entities as nodes in the knowledge graph are instances of entities as nodes in the semantic network. In other words, entities in the semantic network are essentially the entity categories to which entities in the knowledge graph belong.
  • SPO Subject-Predicate-Object
  • One or more embodiments of this specification provide a configuration method, analysis method and device for entities in a knowledge graph.
  • a method for configuring entities in a knowledge graph including: obtaining a logical relational expression of a first entity in the knowledge graph, where the logical relational expression is used to indicate constraint information of the first entity; The logical relational expression is stored in association with the first entity.
  • the logical relational expression includes several operational expressions; adjacent operational expressions are connected through logical operators.
  • the several operational expressions include a first operational expression, used to indicate the second entity that the first entity is allowed to depend on and/or the entity category to which the second entity belongs.
  • the first operation expression includes a plurality of identification information, and a single identification information is used to indicate a second entity that the first entity is allowed to rely on and/or to which the second entity belongs.
  • Entity categories; adjacent identification information is connected through logical operators.
  • the identification information includes the primary key of its corresponding second entity; or the identification information includes the primary key of its corresponding entity category; or the identification information includes the corresponding second entity key.
  • the plurality of operational expressions include a second operational expression, used to indicate a third entity that the first entity is prohibited from relying on and/or an entity category to which the third entity belongs.
  • the plurality of operational expressions include a third operational expression, used to indicate conditions that the attribute information of the second entity on which the first entity depends needs to satisfy.
  • the third operation expression includes several sub-operations; adjacent sub-operations are connected by logical operators.
  • the sub-operation expression includes an attribute name, a relational operator and a condition object arranged in sequence.
  • the relational operator indicates the attribute information corresponding to the attribute name and the condition object. operational relationship between them.
  • the logical relationship is stored in association with the first entity as a semantic rule attribute of the first entity.
  • a method for analyzing entities in a knowledge graph including: determining a first entity to be analyzed in the knowledge graph; reading a logical relationship stored in association with the first entity, where the logical relationship is expressed in Indicating the constraint information of the first entity; using the logical relationship to perform knowledge inference business related to the first entity.
  • the logical relational expression includes several operational expressions; adjacent operational expressions are connected through logical operators.
  • the several operational expressions include a first operational expression, used to indicate the second entity that the first entity is allowed to depend on and/or the entity category to which the second entity belongs.
  • the first operation expression includes a plurality of identification information, and a single identification information is used to indicate a second entity that the first entity is allowed to rely on and/or to which the second entity belongs.
  • Entity categories; adjacent identification information is connected through logical operators.
  • the identification information includes the primary key of its corresponding second entity; or the identification information includes the primary key of its corresponding entity category; or the identification information includes the corresponding second entity key.
  • the plurality of operational expressions include a second operational expression, used to indicate a third entity that the first entity is prohibited from relying on and/or an entity category to which the third entity belongs.
  • the plurality of operational expressions include a third operational expression, used to indicate conditions that the attribute information of the second entity on which the first entity depends needs to satisfy.
  • the third operational expression includes several sub-operational expressions; adjacent sub-operational expressions are connected through logical operators.
  • the sub-operation expression includes an attribute name, a relational operator and a condition object arranged in sequence.
  • the relational operator indicates the attribute information corresponding to the attribute name and the condition object. operational relationship between them.
  • the logical relationship is stored in association with the first entity as a semantic rule attribute of the first entity.
  • a device for configuring entities in a knowledge graph including: a data acquisition unit configured to acquire a logical relational expression of a first entity in the knowledge graph, wherein the logical relational expression is used to indicate the first entity constraint information; a storage processing unit configured to associate and store the logical relational expression with the first entity.
  • an analysis device for entities in a knowledge graph including: an entity determination unit configured to determine the first entity to be analyzed in the knowledge graph; a data reading unit configured to read the information related to the first entity Associated with stored logical relational expressions, the logical relational expressions are used to indicate constraint information of the first entity; an analysis and processing unit is configured to utilize the logical relational expressions to perform knowledge inference services related to the first entity.
  • a computer-readable storage medium is provided, with a computer program/instruction stored thereon.
  • the computer program/instruction is executed in a computing device, the computing device executes any of the above first or second aspects. method described in one item.
  • a computing device including a memory and a processor.
  • Computer programs/instructions are stored in the memory.
  • the processor executes the computer program/instructions, the processor implements the first step described above. The method described in any one of the aspects or the second aspect.
  • the semantic completeness of the knowledge graph is improved by storing the logical relationship expression used to indicate the constraint information of the entity in the knowledge graph in association with the entity.
  • the logical relationship stored in association with the first entity can be used to perform knowledge reasoning services related to the first entity. , which can better implement knowledge reasoning services based on knowledge graphs.
  • Figure 1 is a schematic diagram of a knowledge graph provided as an example in the embodiment of this specification.
  • Figure 2 is a flow chart of a method for configuring entities in a knowledge graph provided in the embodiment of this specification
  • Figure 3 is a flow chart of a method for analyzing entities in a knowledge graph provided in the embodiment of this specification
  • Figure 4 is a schematic diagram of a device for configuring entities in a knowledge graph provided in an embodiment of this specification
  • Figure 5 is a schematic diagram of an analysis device for entities in a knowledge graph provided in an embodiment of this specification.
  • Knowledge graph is a knowledge base that describes entities and relationships between entities in the objective world in the form of graphs. It is a collection of all concepts and their semantic relationships, and is a formal representation of knowledge and knowledge structures. Knowledge graphs usually use a resource description framework to describe data, including nodes, edges and node attributes. It represents and stores instances in the form of triples, establishes semantic links between different instances, and describes objective concepts in a structured form. Entities and their relationships.
  • the knowledge graph can intuitively express entities and the relationships between entities, but it cannot intuitively express the binding rules that establish the relationships between entities; in other words, for a single entity in the knowledge graph, the knowledge graph cannot Express the binding rules that establish the association between the entity and other entities.
  • the binding rules of the entity can also be expressed as the binding information of the entity.
  • Resource description framework (RDFS) language and Web Ontology Language (OWL) can usually define the constraints of an association relationship on the entity types at both ends at the model layer, but cannot describe the same at the instance granularity. Or different instances under different entity categories have different constraint rules. The lack of such constraint rules will result in the semantics of certain relationships not being fully expressed on the knowledge graph, which will result in the inability to directly perform corresponding knowledge inference based on the knowledge graph. business.
  • the relationship "tires are an integral part of a car” can be expressed in a knowledge graph.
  • a similar situation that cannot be expressed is that a car has many components, but some components are necessary to constitute a car. Only having these components at the same time can a car be constituted.
  • Thyroid belongs to the instance of the entity type "examination object”
  • ultrasonic examination and pathological examination belong to the instance of the entity type "examination type”.
  • embodiments of this specification provide a method and device for analyzing entities in a knowledge graph.
  • the semantic completeness of the knowledge graph is improved.
  • the logical relationship stored in association with the first entity can be used to perform knowledge reasoning services related to the first entity. , which can better implement knowledge reasoning services based on knowledge graphs.
  • Figure 2 is a flow chart of a method for configuring entities in a knowledge graph provided in the embodiment of this specification.
  • the method may be performed by any device, device or cluster of devices with computing/processing capabilities. As shown in Figure 2, the method may at least include the following steps 21 and 23.
  • step 21 obtain the logical relationship expression of the first entity in the knowledge graph.
  • the logical relational expression is used to indicate the constraint information of the first entity.
  • the constraint information of any first entity may specifically include but is not limited to at least one of the following information: several entities that the first entity is allowed to rely on, and the entity categories to which each of the several entities that the first entity is allowed to rely on belongs to.
  • Logical relational expressions can include several operational expressions. When there are multiple operation expressions, adjacent operation expressions can be connected through logical operators, where logical operators can include but are not limited to “and", “or", “and”, “or”, etc. Operators that represent logical relationships. For example, the form of the logical relational expression is “expression Op expression”, where "expression” represents the operational expression contained in the logical relational expression, and “Op” represents the logical operator used to connect adjacent operational expressions. It should be noted that "expression” itself can also be nested with several operational expressions "expression” used to represent the existence of logical relationships.
  • a certain logical relationship expression_1 may be equal to expression_2 Op expression_3; in addition, it can also be defined through special symbols
  • the priority of a logical relational expression For example, several logical relational expressions connected in sequence can be placed in arc brackets indicating that they have higher operational priority.
  • the specific form is, for example, expression_4 and (expression_2 or expression_3).
  • the several operational expressions include a first operational expression, used to indicate the second entity that the first entity is allowed to depend on and/or the entity category to which the second entity belongs.
  • the first operational expression can be used to indicate the entities that need to be passed on the communication path ending with the first entity in the knowledge graph and/or the entity category to which the entities that need to be passed belong.
  • the first operation expression includes several identification information, and a single identification information is used to indicate a second entity that the first entity is allowed to rely on and/or the entity category to which the second entity belongs; adjacent identification Information is connected through logical operators.
  • the single identification information may include the primary key/name of its corresponding second entity; or the single identification information may include the primary key/name of its corresponding entity category; or the single identification information may include the primary key/name of its corresponding second entity.
  • the form of the first operation expression is Entity_1 Op Entity_2Op... Entity_n, where Entity_1 to Entity_n are used to represent n pieces of identification information.
  • Entity_i among Entity_1 ⁇ Entity_n, it can indicate the i-th entity that the first entity depends on, for example, Entity_i is the primary key/name of the i-th entity; or it can indicate that the first entity depends on
  • the entity category to which the i-th entity belongs for example, Entity_i is the primary key of the entity category to which the i-th entity belongs; or, it can indicate the i-th entity that the first entity depends on and the entity category to which the i-th entity belongs, for example Entity_i is composed of the primary key of the i-th entity and the primary key of the entity category to which the i-th entity belongs.
  • the primary key of the entity and the primary key of the entity category to which it belongs are separated by a specific separator.
  • the ultrasonic examination characteristics of goiter are related to smooth surface and diffuse enlargement, and its pathological examination characteristics are related to location diffusion.
  • the logical relational expression used to indicate the constraint information of "goiter” may include, for example, operational expressions used to indicate each entity it depends on and the entity category to which each entity it depends belongs: examination object_thyroid and examination characteristics _Diffuse swelling and examination characteristics_Smooth surface and examination characteristics_Location diffusion and examination type_Ultrasound examination and examination type_Pathological examination.
  • the plurality of operational expressions include a second operational expression, used to indicate a third entity that the first entity is prohibited from relying on and/or an entity category to which the third entity belongs.
  • the primary key of the third entity and/or the primary key of the entity category to which the third entity belongs can be used to construct an intermediate expression similar in structure to the aforementioned first operation expression, and then by adding in front of the intermediate expression to represent non- Special symbols of operation such as "! are used to indicate that the first entity is prohibited from relying on the third entity and/or the entity category to which the third entity belongs; alternatively, the primary key of the third entity and/or the entity category to which the third entity belongs can also be used. Add a special symbol such as "! that represents non-operation before the primary key of the entity category, and then use each primary key after adding the special symbol to form a second operation expression with the same structure as the aforementioned first operation expression.
  • the ultrasonic examination results of goiter have nothing to do with morphological rules and clear boundaries, that is, goiter does not depend on the two aforementioned entities.
  • the logical relational expression used to indicate the constraint information of "Goiter” may, for example, include operational expressions used to indicate each entity that it is prohibited to depend on and the entity category to which each entity that it is prohibited to depend on belongs to:! (Check Characteristics_Morphological Rules and Check Characteristics_Clear Boundary); or,! Check Characteristics_Morphological Rules and! Check Features_Clear Boundaries.
  • the plurality of operational expressions include a third operational expression, used to indicate conditions that the attribute information of the second entity on which the first entity depends needs to satisfy.
  • the third operational expression may include several sub-operational expressions, and adjacent sub-operational expressions are connected through logical operators.
  • a single sub-operation may include an attribute name, a relational operator and a condition object arranged in sequence.
  • the form of a single sub-operation is, for example, Entity_Property Op Object or Property Op Object, where Property represents the attribute name, Op represents the relational operator, Object represents the condition object, and Entity represents the primary key of the entity category to which the second entity on which the first entity depends belongs.
  • Object can usually be a numeric value, a string, or an array. The array is only used to support combination with the relational operator "in” to indicate that the value of the attribute information corresponding to the attribute name is located in the array.
  • the entity Zhang San in the knowledge graph belongs to the entity type natural person, and it has an association relationship connected to the entity rural young man belonging to the entity type people.
  • the constraint information of rural young men is not reflected in the knowledge map.
  • rural young men are mainly related to several attributes of natural persons, such as "gender is male, age is between 15 and 30, and household registration type is rural household registration.”
  • the logical relationship expression is associated with the first entity and stored. For example, by extending the semantic rule attribute in the model layer of the first entity, the logical relationship is stored in association with the first entity as the semantic rule attribute of the first entity, so that the semantic rule of the first entity can be obtained in the subsequent process.
  • the rule attribute field reads its associated logical relationship, and uses the logical relationship to perform knowledge inference business related to the first entity.
  • FIG. 3 is a flow chart of a method for analyzing entities in a knowledge graph provided in an embodiment of this specification.
  • the method may be performed by any device, device or cluster of devices with computing/processing capabilities. As shown in Figure 3, the method may at least include the following steps 31 to 35.
  • step 31 the first entity to be analyzed in the knowledge graph is determined.
  • the business data can usually be unstructured data, semi-structured data or structured data.
  • the business data can be pictures, text or natural language processing (Natural Language Processing).
  • entities can be determined from the business data through entity recognition or other data processing methods, so that the identified entities can be used as the first entities to be analyzed in the knowledge map, or they can be associated with the entities in the knowledge map.
  • the entity of the connection path is determined as the first entity to be analyzed.
  • the business data involved may be, for example, a physical examination report related to the thyroid gland, and the first entity may be, for example, the "goiter" in the aforementioned example.
  • step 33 the logical relational expression stored in association with the first entity is read.
  • the logical relational expression is used to indicate the constraint information of the first entity.
  • its associated logical relationship expression can be read from the semantic rule attribute field of the first entity.
  • step 35 the logical relationship is used to perform the knowledge inference business related to the first entity.
  • the knowledge reasoning business may also be a business belonging to other specific technical scenarios.
  • the user's age, gender, household registration type and other attribute information may be combined with the aforementioned
  • the illustratively described entity "young rural male” is associated with a stored logical relationship to match, and based on the matching result it is decided whether the relevant user belongs to the group classification indicated by the entity "young rural male".
  • each information including the first entity in the knowledge graph may be returned to the terminal at the same time.
  • the respective logical relationship expressions of the entities make the respective logical relationship expressions associated with each entity also displayed on the terminal, so that when users query the knowledge graph through the terminal, they can also learn the constraint information of each entity in the knowledge graph.
  • the embodiments of this specification also provide a device for configuring entities in the knowledge graph.
  • the device includes: a data acquisition unit 41 configured to obtain a logical relationship expression of the first entity in the knowledge graph, where the logical relationship expression is used to indicate the constraint information of the first entity; a storage processing unit 43. Configure to associate and store the logical relationship expression with the first entity.
  • the embodiments of this specification also provide an analysis device for entities in a knowledge graph.
  • the device includes: an entity determination unit 51, configured to determine the first entity to be analyzed in the knowledge graph; a data reading unit 53, configured to read the logical relationship expression stored in association with the first entity , the logical relational expression is used to indicate the constraint information of the first entity; the analysis and processing unit 55 is configured to use the logical relational expression to perform knowledge inference services related to the first entity.
  • the functions described in this specification can be implemented using hardware, software, firmware, or any combination thereof.
  • the computer program corresponding to these functions can be stored in a computer-readable medium or transmitted as one or more instructions/codes on the computer-readable medium, so that the computer program corresponding to these functions can be used by the computer.
  • the method described in any embodiment of this specification is implemented through a computer.
  • the embodiments of this specification also provide a computer-readable storage medium on which computer programs/instructions are stored.
  • the computing device executes the information provided in any embodiment of this specification.
  • the embodiments of this specification also provide a computing device, including a memory and a processor.
  • Computer programs/instructions are stored in the memory.
  • the processor executes the computer program/instructions, any embodiment of this specification is implemented. or implement the configuration method of entities in the knowledge graph provided in any embodiment of this specification.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Databases & Information Systems (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Machine Translation (AREA)

Abstract

本说明书实施例中提供了一种知识图谱中实体的分析方法及装置。该方法包括：确定知识图谱中有待分析的第一实体；读取与所述第一实体关联存储的逻辑关系式，所述逻辑关系式用于指示所述第一实体的约束信息；利用所述逻辑关系式执行与所述第一实体相关的知识推理业务。

Description

知识图谱中实体的配置方法、分析方法及装置
本申请要求于2022年03月08日提交中国国家知识产权局、申请号为202210218897.X、申请名称为“知识图谱中实体的配置方法、分析方法及装置”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本说明书一个或多个实施例涉及计算机领域，尤其涉及一种知识图谱中实体的配置方法、分析方法及装置。
背景技术
知识图谱(Knowledge Graph)的建模过程依赖本体知识组成的语义网络，其中本体知识可以利用SPO(Subject-Predicate-Object)三元组来表示，用于描述两个不同实体(概念或对象)间的关系。知识图谱中作为节点的实体是语义网络中作为节点的实体的实例，换而言之即语义网络中的实体实质上是知识图谱中的实体所属的实体类别。
希望有一种新的技术方案，以期有利于更好的实现基于知识图谱的知识推理业务。
发明内容
本说明书一个或多个实施例中提供了一种知识图谱中实体的配置方法、分析方法及装置。
第一方面，提供了一种知识图谱中实体的配置方法，包括：获取知识图谱中第一实体的逻辑关系式，其中所述逻辑关系式用于指示所述第一实体的约束信息；将所述逻辑关系式与所述第一实体关联存储。
在一种可能的实施方式中，所述逻辑关系式中包括若干运算表达式；相邻运算表达式之间通过逻辑运算符连接。
在一种可能的实施方式中，所述若干运算表达式中包括第一运算表达式，用于指示所述第一实体允许依赖的第二实体和/或所述第二实体所属的实体类别。
在一种可能的实施方式中，所述第一运算表达式中包括若干标识信息，单个所述标识信息用于指示所述第一实体允许依赖的一个第二实体和/或该第二实体所属的实体类别；相邻标识信息之间通过逻辑运算符连接。
在一种可能的实施方式中，所述标识信息包括其对应的第二实体的主键；或者，所述标识信息包括其对应的实体类别的主键；或者，所述标识信息包括其对应的第二实体的主键和该第二实体所属的实体类别的主键。
在一种可能的实施方式中，所述若干运算表达式中包括第二运算表达式，用于指示所述第一实体禁止依赖的第三实体和/或所述第三实体所属的实体类别。
在一种可能的实施方式中，所述若干运算表达式中包括第三运算表达式，用于指示所述第一实体依赖的第二实体的属性信息需要满足的条件。
在一种可能的实施方式中，所述第三运算表达式中包括若干子运算式；相邻子运算式 之间通过逻辑运算符连接。
在一种可能的实施方式中，所述子运算式中包括依次排列的属性名称、关系运算符和条件对象，所述关系运算符指示与所述属性名称对应的属性信息和所述条件对象之间的运算关系。
在一种可能的实施方式中，所述逻辑关系式作为所述第一实体的语义规则属性与所述第一实体关联存储。
第二方面，提供了一种知识图谱中实体的分析方法，包括：确定知识图谱中有待分析的第一实体；读取与所述第一实体关联存储的逻辑关系式，所述逻辑关系式用于指示所述第一实体的约束信息；利用所述逻辑关系式执行与所述第一实体相关的知识推理业务。
在一种可能的实施方式中，所述逻辑关系式中包括若干运算表达式；相邻运算表达式之间通过逻辑运算符连接。
在一种可能的实施方式中，所述若干运算表达式中包括第一运算表达式，用于指示所述第一实体允许依赖的第二实体和/或所述第二实体所属的实体类别。
在一种可能的实施方式中，所述第一运算表达式中包括若干标识信息，单个所述标识信息用于指示所述第一实体允许依赖的一个第二实体和/或该第二实体所属的实体类别；相邻标识信息之间通过逻辑运算符连接。
在一种可能的实施方式中，所述标识信息包括其对应的第二实体的主键；或者，所述标识信息包括其对应的实体类别的主键；或者，所述标识信息包括其对应的第二实体的主键和该第二实体所属的实体类别的主键。
在一种可能的实施方式中，所述若干运算表达式中包括第二运算表达式，用于指示所述第一实体禁止依赖的第三实体和/或所述第三实体所属的实体类别。
在一种可能的实施方式中，所述若干运算表达式中包括第三运算表达式，用于指示所述第一实体依赖的第二实体的属性信息需要满足的条件。
在一种可能的实施方式中，所述第三运算表达式中包括若干子运算式；相邻子运算式之间通过逻辑运算符连接。
在一种可能的实施方式中，所述子运算式中包括依次排列的属性名称、关系运算符和条件对象，所述关系运算符指示与所述属性名称对应的属性信息和所述条件对象之间的运算关系。
在一种可能的实施方式中，所述逻辑关系式作为所述第一实体的语义规则属性与所述第一实体关联存储。
第三方面，提供了一种知识图谱中实体的配置装置，包括：数据获取单元，配置为获取知识图谱中第一实体的逻辑关系式，其中所述逻辑关系式用于指示所述第一实体的约束信息；存储处理单元，配置为将所述逻辑关系式与所述第一实体关联存储。
第四方面，提供了一种知识图谱中实体的分析装置，包括：实体确定单元，配置为确定知识图谱中有待分析的第一实体；数据读取单元，配置为读取与所述第一实体关联存储的逻辑关系式，所述逻辑关系式用于指示所述第一实体的约束信息；分析处理单元，配置为利用所述逻辑关系式执行与所述第一实体相关的知识推理业务。
第五方面，提供了一种计算机可读存储介质，其上存储有计算机程序/指令，当所述计算机程序/指令在计算设备中执行时，计算设备执行上述第一方面或第二方面中任一项所述 的方法。
第六方面，提供了一种计算设备，包括存储器和处理器，所述存储器中存储有计算机程序/指令，当所述处理器执行所述计算机程序/指令时，所述处理器实现上述第一方面或第二方面中任一项所述的方法。
通过本说明书一个或多个实施例中提供的方法及装置，通过将用于指示知识图谱中实体的约束信息的逻辑关系式与该实体关联存储，提升了知识图谱的语义完备度。在基于知识图谱执行知识推理业务的过程中，对于知识图谱中任意有待分析的第一实体而言，可以利用与该第一实体关联存储的逻辑关系式执行与该第一实体相关的知识推理业务，从而可以更好的实现基于知识图谱执行知识推理业务。
附图说明
为了更清楚地说明本说明书实施例的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其它的附图。
图1为本说明书实施例中示例性提供的知识图谱的示意图；
图2为本说明书实施例中提供的一种知识图谱中实体的配置方法的流程图；
图3为本说明书实施例中提供的一种知识图谱中实体的分析方法的流程图；
图4为本说明书实施例中提供的一种知识图谱中实体的配置装置的示意图；
图5为本说明书实施例中提供的一种知识图谱中实体的分析装置的示意图。
具体实施方式
下面结合附图，对本说明书所提供的各个非限制性实施例进行详细描述。
知识图谱是以图的形式描述客观世界中实体及实体间关系的知识库，它是所有概念及其语义关系的集合，是知识和知识结构的形式化表征。知识图谱通常使用资源描述框架来描述数据，包括节点、边和节点的属性，其以三元组的形式表示和存储实例，建立不同实例之间的语义链接，并以结构化形式描述客观概念、实体及其关系。
知识图谱能够很直观地表达实体和实体间的关联关系，然而其无法直观表达出实体间的关联关系成立的约束性规则；换而言之即对于知识图谱中的单个实体而言，知识图谱无法表达该实体与其它实体间的关联关系成立的约束性规则，其中该实体的约束性规则也可表述为该实体的约束信息。资源描述框架(resource description framework，RDFS)语言和网络本体语言(Web Ontology Language，OWL)等通常能够在模型层定义某个关联关系对于其两端实体类型的约束，无法在实例这个粒度上描述相同或不同实体类别下的不同实例具有不同的约束规则，而缺失这样的约束规则将会导致某些关联关系的语义无法在知识图谱上进行完整表达，进而导致无法基于知识图谱直接的进行相应知识推理业务。
举例来说，知识图谱中可以表达“轮胎是轿车的组成部分”这个关系，我们能够通过这个关系知道轮胎是轿车的组成部分，但轿车必须有4个轮胎这样的语义约束就无法通过知识图谱表达。类似无法表达的情况还有比如轿车有很多的组成部分，但有些组成部分是构成轿车的必要条件，只有同时拥有这些组成部分才能构成轿车。更具体地，请参见图1示 例性提供的用于核保或核赔业务的知识图谱，甲状腺属于实体类型“检查对象”的实例，超声检查和病理检查属于实体类型“检查类型”的实例，形态规则、弥漫性肿大、表面光滑、边界清晰和位置扩散均是属于实体类型“检查特征”的实例，甲状腺肿是属于实体类型“病症”的实例；甲状腺肿的超声检查特征与表面光滑和弥漫性肿大相关，且其病例检查特征与位置扩散相关，然而在知识图谱中并未体现甲状腺肿依赖的实体是“弥漫性肿大”、“表面光滑”和“位置扩散”而并不依赖其它实体，导致无法基于知识图谱直接执行相关于“甲状腺肿”的知识推理业务。
鉴于以上问题，本说明书实施例中提供了一种知识图谱中实体的分析方法及装置。通过将用于指示知识图谱中实体的约束信息的逻辑关系式与该实体关联存储，提升了知识图谱的语义完备度。在基于知识图谱执行知识推理业务的过程中，对于知识图谱中任意有待分析的第一实体而言，可以利用与该第一实体关联存储的逻辑关系式执行与该第一实体相关的知识推理业务，从而可以更好的实现基于知识图谱执行知识推理业务。
图2为本说明书实施例中提供的一种知识图谱中实体的配置方法的流程图。该方法可以由任何具有计算/处理能力的装置、设备或设备集群执行。如图2所示，该方法至少可以包括如下步骤21和步骤23。
首先，在步骤21，获取知识图谱中第一实体的逻辑关系式。其中，该逻辑关系式用于指示第一实体的约束信息。
对于知识图谱中的不同实体，其通常可以具有不同的约束信息；约束信息可以由用户提供，更具体地说逻辑关系式可以由用户直接提供或者通过对用户提供的表征约束信息的文本数据进行语义分析以生成。对于任意的第一实体的约束信息，具体可以包括但不限于如下各项信息中的至少一项：该第一实体允许依赖的若干实体、该第一实体允许依赖的若干实体各自所属的实体类别、该第一实体禁止依赖的若干实体、该第一实体禁止依赖的若干实体各自所属的实体类别、该第一实体依赖的若干实体各自的属性信息需要满足的条件。
逻辑关系式中可以包括若干运算表达式。当存在多个运算表达式的情况下，相邻运算表达式之间可以通过逻辑运算符连接，其中逻辑运算符可以包括但不限于“和”、“或”、“and”、“or”等表征逻辑关系的运算符。示例性的，逻辑关系式的形态例如为“expression Op expression”，其中“expression”表征该逻辑关系式所包含的运算表达式，“Op”表征用于连接相邻运算表达式的逻辑运算符。需要特别说明的是，“expression”自身还可以嵌套若干用于表征存在逻辑关系的运算表达式“expression”，例如某个逻辑关系式expression_1可能相等于expression_2 Op expression_3；此外还可以通过特殊符号定义逻辑关系式的优先级，例如依次连接的若干逻辑关系式可以位于表征其具有更高运算优先级的圆弧括号中，具体形态例如为expression_4 and(expression_2 or expression_3)。
在一种可能的实施方式中，若干运算表达式中包括第一运算表达式，用于指示第一实体允许依赖的第二实体和/或该第二实体所属的实体类别。换而言之该第一运算表达式可以用于指示知识图谱中以第一实体为终点的联通路径上需要经过的实体和/或需要经过的实体所属的实体类别。在一个较为具体的示例中，第一运算表达式中包括若干标识信息，单个标识信息用于指示第一实体允许依赖的一个第二实体和/或该第二实体所属的实体类别；相邻标识信息之间通过逻辑运算符连接。更具体地说，单个标识信息可以包括其对应的第二实体的主键/名称；或者，单个标识信息包括其对应的实体类别的主键/名称；或者，单 个标识信息包括其对应的第二实体的主键/名称和该第二实体所属的实体类别的主键/名称。
举例来说，第一运算表达式的形态例如为Entity_1 Op Entity_2Op……Entity_n，其中Entity_1～Entity_n用于表征n个标识信息。对于Entity_1～Entity_n中任意的第i个标识信息Entity_i而言，其可以指示第一实体依赖的第i个实体，例如Entity_i为第i个实体的主键/名称；或者，其可以指示第一实体依赖的第i个实体所属的实体类别，例如Entity_i为第i个实体所属的实体类别的主键；或者，其可以指示第一实体依赖的第i个实体以及该第i个实体所属的实体类别，例如Entity_i由第i个实体的主键和第i个实体所属的实体类别的主键拼接而成，其中实体的主键和其所属实体类别的主键之间通过特定分隔符分隔。
以前述图1示例的知识图谱为例，甲状腺肿的超声检查特征与表面光滑和弥漫性肿大相关，且其病理检查特征与位置扩散相关。那么用于指示“甲状腺肿”的约束信息的逻辑关系式，例如可以包括用于指示其依赖的各个实体以及其依赖的各个实体各自所属的实体类别的运算表达式：检查对象_甲状腺and检查特征_弥漫性肿大and检查特征_表面光滑and检查特征_位置扩散and检查类型_超声检查and检查类型_病理检查。
在一种可能的实施方式中，若干运算表达式中包括第二运算表达式，用于指示第一实体禁止依赖的第三实体和/或该第三实体所属的实体类别。举例来说，可利用第三实体的主键和/或第三实体所属的实体类别的主键，构建与前述第一运算表达式结构相似的中间表达式，然后通过在该中间表达式前增加表征非运算的特殊符号例如“！”来实现指示第一实体禁止依赖的第三实体和/或该第三实体所属的实体类别；或者，还可以通过在第三实体的主键和/或第三实体所属的实体类别的主键前增加表征非运算的特殊符号例如“！”，进而利用增加特殊符号后的各个主键构成与前述第一运算表达式结构相同的第二运算表达式。
以前述图1示例的知识图谱为例，甲状腺肿的超声检查结果与形态规则和边界清晰无关，即甲状腺肿并不依赖于前述两个实体。那么用于指示“甲状腺肿”的约束信息的逻辑关系式，例如可以包括用于指示其禁止依赖的各个实体以及其禁止依赖的各个实体各自所属的实体类别的运算表达式：！(检查特征_形态规则and检查特征_边界清晰)；或者，！检查特征_形态规则and！检查特征_边界清晰。
在一种可能的实施方式中，若干运算表达式中包括第三运算表达式，用于指示第一实体依赖的第二实体的属性信息需要满足的条件。其中该第三运算表达式中可以包括若干子运算式，相邻子运算式之间通过逻辑运算符连接。
单个子运算式中可以包括依次排列的属性名称、关系运算符和条件对象，关系运算符指示与属性名称对应的属性信息和条件对象之间的运算关系。关系运算符具体例如为“in”、“>”、“>＝”、“<”、“<＝”、“＝＝”、“！＝”或者“contains”等等。单个子运算式的形态例如为Entity_Property Op Object或者Property Op Object，其中Property表征属性名称、Op表征关系运算符、Object表征条件对象，Entity表征第一实体依赖的第二实体所属的实体类别的主键。Object通常可以为数值、字符串或数组，其中数组仅用于支持与关系运算符“in”进行组合以表示与属性名称对应的属性信息的取值位于该数组中。
举例来说，知识图谱中的实体张三属于实体类型自然人，其存在关联关系连接到属于实体类型人群的实体农村男青年。农村男青年的约束信息并没有在知识图谱中体现，实际上农村男青年主要跟自然人的几个属性相关，比如“性别是男性，年龄在15到30之间，户口类型是农村户口”。那么用于指示农村男青年的约束信息的逻辑关系式，例如可以包括第 三运算表达式：自然人_性别＝＝"男性"and(自然人_年龄>＝15and自然人_年龄<＝30)and自然人_户口类型＝＝"农村"。
接着，在步骤23，将逻辑关系式与第一实体关联存储。例如通过在该第一实体的模型层扩展语义规则(semantic rule)属性，该逻辑关系式作为第一实体的语义规则属性与第一实体关联存储，以便后续过程中可以从该第一实体的语义规则属性字段读取其关联的逻辑关系式，并利用该逻辑关系式执行相关于第一实体的知识推理业务。
具体地，请参照图3，图3为本说明书实施例中提供的一种知识图谱中实体的分析方法的流程图。该方法可以由任何具有计算/处理能力的装置、设备或设备集群执行。如图3所示，该方法至少可以包括如下步骤31～步骤35。
首先，在步骤31，确定知识图谱中有待分析的第一实体。
用户可通过终端发起包含业务数据的查询请求，该业务数据通常可以是非结构化数据、半结构化数据或者结构化数据，例如该业务数据可以是图片、文本或者经过自然语言处理(Natural Language Processing，NLP)得到的具有特定格式的表格或文档。可以通过实体识别或者其它数据处理方式从该业务数据中确定出若干实体，从而可以在知识图谱中将确定的若干实体分别作为有待分析的第一实体，或者在知识图谱中将与该若干实体存在联通路径的实体确定为有待分析的第一实体。以前述的核保业务或核赔业务为例，其涉及的业务数据例如可以为与甲状腺相关的体检报告，第一实体例如为前述示例的“甲状腺肿”。
接着，在步骤33，读取与第一实体关联存储的逻辑关系式。其中，该逻辑关系式用于指示第一实体的约束信息，其具体结构可参照前文的对逻辑关系式的相关描述，这里不再赘述。步骤23中具体可以从该第一实体的语义规则属性字段读取其关联的逻辑关系式。
最后，在步骤35，利用逻辑关系式执行与第一实体相关的知识推理业务。
以前述示例的核保业务或核赔业务为例，其在执行过程中可能需要基于图1所示的知识图谱和用户的体检报告，推理用户是否存在疾病甲状腺肿，该过程中则涉及相关于实体甲状腺肿的知识推理。此种情况下，可以通过对体检报告进行实体识别及其它处理，提取体检报告中包含的若干实体、该若干实体各自所属的实体类型以及各自的属性信息等等，通过将从体检报告中提取的前述各项信息与甲状腺肿关联存储的逻辑关系式进行匹配，如果匹配则可推理出用户可能患有疾病甲状腺肿，后续可能从知识图谱中抽取包含甲状腺肿的联通路径，决策是否允许用户参保或是否允许对用户理赔等。
虽然前面已经对知识推理业务进行了示例性描述，然而可以理解的是知识推理业务具体还可能为属于其它具体技术场景中的业务，例如可能将用户的年龄、性别、户口类型等属性信息与前述示例性描述的实体“农村男青年的”关联存储的逻辑关系式进行匹配，基于匹配结果决策相关用户是否属于实体“农村男青年”指示的人群分类。
在一些实施例中，当响应于用户通过终端对知识图谱发起的查询请求，向终端返回相应的信息以在终端显示知识图谱时，可以同时向终端返回知识图谱中包含第一实体在内的各个实体各自的逻辑关系式，使得每个实体各自关联的逻辑关系式也在终端进行显示，从而使得用户在通过终端查询知识图谱时，同时了解到知识图谱中各实体的约束信息。
与前述方法实施例基于相同的构思，本说明书实施例中还提供了一种知识图谱中实体的配置装置。如图4所示，该装置包括：数据获取单元41，配置为获取知识图谱中第一实体的逻辑关系式，其中所述逻辑关系式用于指示所述第一实体的约束信息；存储处理单元 43，配置为将所述逻辑关系式与所述第一实体关联存储。
与前述方法实施例基于相同的构思，本说明书实施例中还提供了一种知识图谱中实体的分析装置。如图5所示，该装置包括：实体确定单元51，配置为确定知识图谱中有待分析的第一实体；数据读取单元53，配置为读取与所述第一实体关联存储的逻辑关系式，所述逻辑关系式用于指示所述第一实体的约束信息；分析处理单元55，配置为利用所述逻辑关系式执行与所述第一实体相关的知识推理业务。
本领域技术人员应该可以意识到，在上述一个或多个示例中，本说明书所描述的功能可以用硬件、软件、固件或它们的任意组合来实现。当使用软件实现时，可以将这些功能所对应的计算机程序存储在计算机可读介质中或者作为计算机可读介质上的一个或多个指令/代码进行传输，以便这些功能所对应的计算机程序被计算机执行时，通过计算机实现本说明书任意一个实施例中所述的方法。
本说明书实施例中还提供了一种计算机可读存储介质，其上存储有计算机程序/指令，当所述计算机程序/指令在计算设备中执行时，计算设备执行本说明书任意一个实施例中提供的知识图谱中实体的分析方法，或者实现本说明书任意一个实施例中提供的知识图谱中实体的配置方法。
本说明书实施例中还提供了一种计算设备，包括存储器和处理器，所述存储器中存储有计算机程序/指令，所述处理器执行所述计算机程序/指令时，实现本说明书任意一个实施例中提供的知识图谱中实体的分析方法，或者实现本说明书任意一个实施例中提供的知识图谱中实体的配置方法。
本说明书中的各个实施例均采用递进的方式描述，各个实施例中相同、相似的部分互相参见即可，每个实施例中重点说明的都是与其他实施例的不同之处。尤其，对于装置实施例而言，由于其基本相似于方法实施例，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。
上述对本说明书特定实施例进行了描述。其它实施例在所附权利要求书的范围内。在一些情况下，在权利要求书中记载的动作或步骤可以按照不同于实施例中的顺序来执行并且仍然可以实现期望的结果。另外，在附图中描绘的过程不一定要求示出的特定顺序或者连续顺序才能实现期望的结果。在某些实施方式中，多任务处理和并行处理也是可以的或者可能是有利的。
以上所述的具体实施方式，对本发明的目的、技术方案和有益效果进行了进一步详细说明，所应理解的是，以上所述仅为本发明的具体实施方式而已，并不用于限定本发明的保护范围，凡在本发明的技术方案的基础之上，所做的任何修改、等同替换、改进等，均应包括在本发明的保护范围之内。

Claims (15)

  1. 一种知识图谱中实体的配置方法，包括：
    获取知识图谱中第一实体的逻辑关系式，其中所述逻辑关系式用于指示所述第一实体的约束信息；
    将所述逻辑关系式与所述第一实体关联存储。
  2. 一种知识图谱中实体的分析方法，包括：
    确定知识图谱中有待分析的第一实体；
    读取与所述第一实体关联存储的逻辑关系式，所述逻辑关系式用于指示所述第一实体的约束信息；
    利用所述逻辑关系式执行与所述第一实体相关的知识推理业务。
  3. 根据权利要求1或2所述的方法，其中，所述逻辑关系式中包括若干运算表达式；相邻运算表达式之间通过逻辑运算符连接。
  4. 根据权利要求3所述的方法，其中，所述若干运算表达式中包括第一运算表达式，用于指示所述第一实体允许依赖的第二实体和/或所述第二实体所属的实体类别。
  5. 根据权利要求4所述的方法，其中，所述第一运算表达式中包括若干标识信息，单个所述标识信息用于指示所述第一实体允许依赖的一个第二实体和/或该第二实体所属的实体类别；相邻标识信息之间通过逻辑运算符连接。
  6. 根据权利要求5所述的方法，其中，所述标识信息包括其对应的第二实体的主键；或者，所述标识信息包括其对应的实体类别的主键；或者，所述标识信息包括其对应的第二实体的主键和该第二实体所属的实体类别的主键。
  7. 根据权利要求3所述的方法，其中，所述若干运算表达式中包括第二运算表达式，用于指示所述第一实体禁止依赖的第三实体和/或所述第三实体所属的实体类别。
  8. 根据权利要求3所述的方法，其中，所述若干运算表达式中包括第三运算表达式，用于指示所述第一实体依赖的第二实体的属性信息需要满足的条件。
  9. 根据权利要求8所述的方法，其中，所述第三运算表达式中包括若干子运算式；相邻子运算式之间通过逻辑运算符连接。
  10. 根据权利要求9所述的方法，其中，所述子运算式中包括依次排列的属性名称、关系运算符和条件对象，所述关系运算符指示与所述属性名称对应的属性信息和所述条件对象之间的运算关系。
  11. 根据权利要求1或2所述的方法，其中，所述逻辑关系式作为所述第一实体的语义规则属性与所述第一实体关联存储。
  12. 一种知识图谱中实体的配置装置，包括：
    数据获取单元，配置为获取知识图谱中第一实体的逻辑关系式，其中所述逻辑关系式用于指示所述第一实体的约束信息；
    存储处理单元，配置为将所述逻辑关系式与所述第一实体关联存储。
  13. 一种知识图谱中实体的分析装置，包括：
    实体确定单元，配置为确定知识图谱中有待分析的第一实体；
    数据读取单元，配置为读取与所述第一实体关联存储的逻辑关系式，所述逻辑关系式 用于指示所述第一实体的约束信息；
    分析处理单元，配置为利用所述逻辑关系式执行与所述第一实体相关的知识推理业务。
  14. 一种计算机可读存储介质，其上存储有计算机程序，当所述计算机程序在计算设备中执行时，计算设备执行权利要求1-11中任一项所述的方法。
  15. 一种计算设备，包括存储器和处理器，所述存储器中存储有计算机程序，所述处理器执行所述计算机程序时，实现权利要求1-11中任一项所述的方法。
PCT/CN2023/070481 2022-03-08 2023-01-04 知识图谱中实体的配置方法、分析方法及装置 Ceased WO2023169072A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210218897.X 2022-03-08
CN202210218897.XA CN114297411B (zh) 2022-03-08 2022-03-08 知识图谱中实体的配置方法、分析方法及装置

Publications (1)

Publication Number Publication Date
WO2023169072A1 true WO2023169072A1 (zh) 2023-09-14

Family

ID=80978548

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/070481 Ceased WO2023169072A1 (zh) 2022-03-08 2023-01-04 知识图谱中实体的配置方法、分析方法及装置

Country Status (2)

Country Link
CN (1) CN114297411B (zh)
WO (1) WO2023169072A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117494806A (zh) * 2023-12-28 2024-02-02 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) 基于知识图谱和大语言模型的关系抽取方法、系统及介质
CN118014739A (zh) * 2024-02-18 2024-05-10 德联易控科技(北京)有限公司 一种基于双层图谱的理赔决策规则生成方法及装置
CN120508317A (zh) * 2025-07-22 2025-08-19 苏州元脑智能科技有限公司 指令执行方法、装置、电子设备和存储介质
CN121255776A (zh) * 2025-12-03 2026-01-02 蓝卓数字科技有限公司 一种基于uns的多元异构数据建模方法、系统及相关设备

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114297411B (zh) * 2022-03-08 2023-05-02 支付宝(杭州)信息技术有限公司 知识图谱中实体的配置方法、分析方法及装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018226404A1 (en) * 2017-06-09 2018-12-13 Microsoft Technology Licensing, Llc Machine reasoning based on knowledge graph
CN109508381A (zh) * 2018-09-29 2019-03-22 北京国双科技有限公司 知识图谱的处理方法及装置
CN112732940A (zh) * 2021-01-15 2021-04-30 医渡云（北京）技术有限公司 基于模型的医学知识图谱的推理方法、装置、设备及介质
CN113868433A (zh) * 2021-09-24 2021-12-31 联想(北京)有限公司 数据处理方法、装置及电子设备
CN114297411A (zh) * 2022-03-08 2022-04-08 支付宝(杭州)信息技术有限公司 知识图谱中实体的配置方法、分析方法及装置

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3055786A4 (en) * 2013-10-09 2017-05-17 Google, Inc. Automatic definition of entity collections
CN110119292B (zh) * 2018-02-07 2023-02-07 中兴通讯股份有限公司 系统运行参数查询方法、匹配方法、装置及节点设备
US11321304B2 (en) * 2019-09-27 2022-05-03 International Business Machines Corporation Domain aware explainable anomaly and drift detection for multi-variate raw data using a constraint repository
CN111950285B (zh) * 2020-07-31 2024-01-23 合肥工业大学 多模态数据融合的医疗知识图谱智能自动构建系统和方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018226404A1 (en) * 2017-06-09 2018-12-13 Microsoft Technology Licensing, Llc Machine reasoning based on knowledge graph
CN109508381A (zh) * 2018-09-29 2019-03-22 北京国双科技有限公司 知识图谱的处理方法及装置
CN112732940A (zh) * 2021-01-15 2021-04-30 医渡云（北京）技术有限公司 基于模型的医学知识图谱的推理方法、装置、设备及介质
CN113868433A (zh) * 2021-09-24 2021-12-31 联想(北京)有限公司 数据处理方法、装置及电子设备
CN114297411A (zh) * 2022-03-08 2022-04-08 支付宝(杭州)信息技术有限公司 知识图谱中实体的配置方法、分析方法及装置

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117494806A (zh) * 2023-12-28 2024-02-02 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) 基于知识图谱和大语言模型的关系抽取方法、系统及介质
CN117494806B (zh) * 2023-12-28 2024-03-08 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) 基于知识图谱和大语言模型的关系抽取方法、系统及介质
CN118014739A (zh) * 2024-02-18 2024-05-10 德联易控科技(北京)有限公司 一种基于双层图谱的理赔决策规则生成方法及装置
CN120508317A (zh) * 2025-07-22 2025-08-19 苏州元脑智能科技有限公司 指令执行方法、装置、电子设备和存储介质
CN121255776A (zh) * 2025-12-03 2026-01-02 蓝卓数字科技有限公司 一种基于uns的多元异构数据建模方法、系统及相关设备

Also Published As

Publication number Publication date
CN114297411A (zh) 2022-04-08
CN114297411B (zh) 2023-05-02

Similar Documents

Publication Publication Date Title
US11763175B2 (en) Systems and methods for semantic inference and reasoning
WO2023169072A1 (zh) 知识图谱中实体的配置方法、分析方法及装置
US10877979B2 (en) Determining explanations for predicted links in knowledge graphs
Shvaiko et al. Ontology matching: state of the art and future challenges
Spanos et al. Bringing relational databases into the semantic web: A survey
JP6693252B2 (ja) 類似性計算装置、薬の類似性を計算し及び類似性を用いて副作用を推定する副作用決定装置及びシステム
US10157350B2 (en) Context based conversation system
US10614086B2 (en) Orchestrated hydration of a knowledge graph
WO2020139861A1 (en) Constructing a knowledge graph employing multiple subgraphs and a linking layer including multiple linking nodes
WO2023029506A1 (zh) 病情分析方法、装置、电子设备及存储介质
CN107391677A (zh) 携带实体关系属性的中文通用知识图谱的生成方法及装置
JP2016100005A (ja) リコンサイル方法、プロセッサ及び記憶媒体
CN116821301A (zh) 基于知识图谱的问题应答方法、装置、介质及计算机设备
Rastogi et al. Personal health knowledge graphs for patients
US11113300B2 (en) System and method for enabling interoperability between a first knowledge base and a second knowledge base
Ferranti et al. Formalizing and validating Wikidata’s property constraints using SHACL and SPARQL
Benedetti et al. Exposing the underlying schema of LOD sources
US20200294666A1 (en) Adding new electronic events into an electronic user profile using a language-independent data format
Achouri et al. Smart fog computing for efficient situations management in smart health environments
Mahfoud Expressive top-k matching for conditional graph patterns
Vaccari et al. An evaluation of ontology matching in geo-service applications
CN114461815A (zh) 家谱知识图谱的噪音检测方法及其装置、电子设备
CN114201602B (zh) 为可视化数据分析生成自然语言语句的方法及装置
US20150370783A1 (en) Method of using a semantic web data source in a target application
Suhas et al. Talking buildings: Interactive human-building smart-bot for smart buildings

Legal Events

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

Ref document number: 23765652

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 23765652

Country of ref document: EP

Kind code of ref document: A1