CN115757810A - Method for constructing standard ontology of knowledge graph - Google Patents
Method for constructing standard ontology of knowledge graph Download PDFInfo
- Publication number
- CN115757810A CN115757810A CN202211298266.XA CN202211298266A CN115757810A CN 115757810 A CN115757810 A CN 115757810A CN 202211298266 A CN202211298266 A CN 202211298266A CN 115757810 A CN115757810 A CN 115757810A
- Authority
- CN
- China
- Prior art keywords
- knowledge
- standard
- ontology
- data
- graph
- 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.)
- Pending
Links
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a method for constructing a standard ontology of a knowledge graph, which belongs to the technical field of power grid knowledge graph construction, and comprises the following steps: step 1: and (3) information processing: the transformer and circuit breaker standard specification files in the multi-source database are processed, and the method comprises the following steps: extracting information, namely extracting multi-source and heterogeneous data, indexing, producing and integrating to complete fragmentation, serialization and semantization of the data; information fusion, integrating the extracted information; processing information, namely classifying and storing fragment knowledge generated by the fused information, and constructing a standard knowledge base of the transformer and the breaker after quality evaluation; step 2: constructing an ontology: the method and the tool for knowledge organization, storage, extraction and reasoning are novel, automatic and intelligent, massive discrete information points are aggregated into a semantic network, and a mature and stable map construction functional system in the industry is introduced to enable the link to be twice with little effort.
Description
Technical Field
The invention relates to the technical field of power grid knowledge graph construction, in particular to a method for constructing a knowledge graph standard ontology.
Background
With the development of science and technology, the traditional knowledge organization and management mode cannot meet the requirements of the current power system. Currently, knowledge bases based on knowledge representation and knowledge reasoning are used in power systems, such as: and an intelligent decision system, a fault positioning system, a power transmission network planning decision and the like are combined with the traditional expert system.
However, most of these knowledge bases rely on the traditional knowledge management method of extracting, sorting and storing data in the form of icons in the database by experts, the knowledge structure that can be stored is single, and each update requires a great deal of time for professional technicians. Particularly for the fields with rapid knowledge change, such as power dispatching, equipment management, data interaction, service inquiry and the like, the existing knowledge management mode is seriously lagged behind the development requirement of the system.
Disclosure of Invention
In order to make up for the defects, the invention provides a method for constructing a knowledge graph standard ontology, and aims to develop the key technology application research of a standard knowledge production platform (knowledge base). In order to research and develop related key technologies around the processes of standard structuring, fragmenting, indexing, modeling, knowledge metaplasia, mapping, intellectualization and the like in the field of standard digital transformation, a company standard knowledge base and a main network equipment knowledge map are constructed.
The invention is realized in the following way: a method for constructing a standard ontology of a knowledge graph comprises the following steps:
step 1: and (3) information processing: the transformer and circuit breaker standard specification files in the multi-source database are processed, and the method comprises the following steps:
extracting information, namely extracting multi-source data and heterogeneous data, indexing, producing and integrating the data, completing fragmentation, serialization and semantization of the data, extracting or learning entities, attributes and interrelations among the entities from a multi-source database, and forming ontology information expression;
information fusion, integrating the extracted information to eliminate contradiction and ambiguity and produce fragment types including seal bars, terms, indexes, formulas, pictures, tables and annexes;
processing information, namely classifying and storing fragment knowledge generated by the fused information, and constructing a standard knowledge base of the transformer and the breaker after quality evaluation;
and 2, step: constructing an ontology: constructing a mapping layer for representing the mapping relation between the meta-ontology and the ontology in the standard data based on the knowledge base obtained in the step 1, wherein the meta-ontology is a common and substantial feature knowledge base extracted from a plurality of ontologies and is used for carrying out abstract expression on the ontology, determining a meta-ontology model and a mapping function between the meta-ontology and the ontology, correspondingly linking each meta-ontology and each ontology according to the mapping function, and constructing the mapping layer;
the determining a meta-ontology model and a mapping function between a meta-ontology and an ontology includes: determining a meta-ontology model; determining a first mapping function for characterizing the mapping of meta-ontologies to ontologies according to the meta-ontology model; correspondingly linking each element body and each body according to the mapping function to construct the mapping layer; constructing a standard knowledge graph based on the mapping layer, wherein the standard knowledge graph comprises the following steps: constructing an ontology model based on the mapping layer; and extracting corresponding entities from the standard text according to the ontology model, and constructing a knowledge graph standard ontology.
In a preferred technical scheme of the invention, in the step 1, not less than 1600 standard specifications of transformers and circuit breakers are processed, including multisource, heterogeneous data extraction, indexing and production, and integration, so as to complete fragmentation, serialization and semantization of the transformer and circuit breakers, the types of the generated fragments need to include seal bars, terms, indexes, formulas, pictures, tables, addendums and the like, and the processing accuracy is required to be more than 95%; the generated fragment knowledge is classified and stored, a standard knowledge base of the transformer and the breaker is constructed, functions of data warehousing, resource management, resource safety management, version management, system management, data storage, backup and the like are supported, and a shared service capability can be formed and is open to the outside.
In a preferred technical scheme of the invention, the standard specifications of the transformer and the breaker comprise national standards, enterprise standards, industry standards, group standards, technical specifications, operation instruction books, anti-accident measures and typical designs.
In a preferred technical scheme of the invention, in step 1, the processing standard specification of the transformer and the breaker equipment is as follows:
a. chapter bar: extracting and processing the seal strip of each level, supporting the seal strip to be nested in a correlated manner, namely processing in a hierarchical manner, wherein the seal strip of the parent level can contain the seal strip of the child level;
b. the terms: the method supports the extraction of terms, and the extraction result comprises term names, term definitions and the like;
c. index and index value: extracting indexes in the standard and outputting the indexes in a key-value form;
d. the formula: extracting formulas in the standard, wherein the formulas need to contain formula names and specific formulas;
e. picture: extracting non-pure character pictures in the standard, and outputting picture resources which need to include picture names and picture resource files;
f. table (b): extracting the table in the standard, wherein the extraction result supports extraction of two types, namely pictures and excel tables, including table names, table headers, row data, column data and the like;
in a preferred technical scheme of the invention, in step 1, a transformer and circuit breaker equipment technical standard knowledge base model is constructed:
a. the method comprises the steps of constructing a technical standard knowledge base of the transformer and the breaker equipment, storing information data into a digital resource base from the processing and warehousing of resources, managing metadata, digital objects, XML data and the like of the resources, and constructing a technical standard knowledge base of a main network transformer;
b. the system has the functions of data warehousing, resource management, resource safety management, version management, system management, data storage, backup and the like, and the abstract sharing service capability is open to the outside;
c. the technical standard knowledge base of the transformer and the breaker equipment needs to comprise a plurality of sub-bases such as a standard file sub-base, a term sub-base, a Zhang Tiao sub-base, an index sub-base, a picture sub-base, a table sub-base and a formula sub-base, wherein each sub-base needs to support data addition, deletion, modification and query, and a front-end graphical page is needed for a user to operate.
In a preferred technical scheme of the invention, in the step 2, the constructed knowledge graph standard ontology comprises the following functions of extracting the standard knowledge of the transformer and the breaker equipment and constructing a graph: knowledge extraction provides knowledge extraction services aiming at different data sources, and all the knowledge extraction services run periodically in the background in a task mode, so that continuous access of various external data is guaranteed. And the automatic construction of the source database data to the knowledge graph is completed through structured and unstructured data access, and the structured data graph entering capability is provided.
In a preferred embodiment of the present invention, the structured data mapping capability includes:
a) And (3) displaying the standard knowledge graph of the transformer and the breaker equipment: visual map data including entity attributes, relationship query between entities, entity attribute query, and the like; classifying and counting knowledge graph data and visually managing graph data content; supporting upper layer application, providing an interface for inquiring entities, attributes and relationships; the method supports not less than two modeling modes such as lists, visual graphs and the like, and supports graphical entities, relationships and attribute editing;
b) And (3) standard knowledge graph construction management of the transformer and the breaker equipment: supporting user authority distribution management; visual management map storage is supported; manual intervention or automatic extraction and increase of atlas data (including schema), visual management and traceability of historical atlases are supported;
c) The standard multi-mode knowledge understanding of the transformer and the breaker equipment is as follows: the method supports extraction of knowledge from documents such as PDF, WORD and TXT, and constructs a knowledge graph; aiming at different data forms, text representation information of structural features is used as an analysis object, and text feature calculation and text feature selection are carried out on material contents by utilizing mature technical methods in the fields of machine learning, natural language processing speech recognition, deep learning and the like, combining field problems and actual experience and combining a related database;
d) Other functional requirements are as follows: model training for extracting types such as entities, attributes and the like is supported; supporting the condition of a visual display platform, a training process and result evaluation; providing basic word segmentation and entity recognition capabilities; supporting the corpus labeling capability, enabling a user to customize a label and supporting the labeling of multimode data; the method has a complete knowledge map construction platformization function, and has full stack construction capabilities of knowledge representation, knowledge modeling, knowledge extraction, knowledge fusion, knowledge storage, knowledge calculation and the like; the system has a complete knowledge application platform function and knowledge application capabilities of knowledge retrieval, knowledge question answering and online relation reasoning based on the map.
In a preferred technical solution of the present invention, in step 2, the constructed knowledge graph standard ontology further includes a graph storage and query function: the storage, processing and data synchronous updating of structured, semi-structured and other data sources in the process of map construction are supported; supporting relationship management among entities, including adding and deleting edge relationships, setting a plurality of relationship objects and the like; the entity retrieval, entity relation calculation, feature query service and the like of the knowledge graph can be realized; and the knowledge graph content is retrieved and displayed through a complete standard graph query statement.
The invention has the beneficial effects that: the whole project starts from the current situation and the requirement of standard digital transformation, namely digital standard construction, top-level design is firstly carried out, then standard specification formulation of digitization, fragmentation and indexing is carried out, meanwhile, research on core key technologies is carried out, including standard document digitization related technology, data processing indexing technology, intelligent service technology and the like, then tool integration and development are carried out, a standard knowledge base of digitization and knowledge metaplasia is realized, finally, an intelligent application platform is built, and scene services are provided for business application. The novel, automatic and intelligent knowledge organization, storage, extraction and reasoning method and tool can aggregate massive discrete information points into a semantic network, and introduce a graph construction functional system mature and stable in the industry to enable the link to be double in result.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
FIG. 1 is a flow chart of a method for constructing a standard ontology of a knowledge graph according to an embodiment of the invention;
FIG. 2 is a schematic structural diagram of a knowledge-graph standard ontology construction subsystem provided by an embodiment of the present invention;
FIG. 3 is a schematic diagram of a knowledge production subsystem according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a knowledge-graph question-answering system according to an embodiment of the present invention;
fig. 5 is a schematic diagram of a knowledge production system of a knowledge center according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings of the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
Examples
Referring to fig. 1, the present invention provides a technical solution: a method for constructing a standard ontology of a knowledge graph comprises the following steps:
step 1: and (3) information processing: the transformer and circuit breaker standard specification files in the multi-source database are processed, and the method comprises the following steps:
extracting information, namely extracting multi-source data and heterogeneous data, indexing, producing and integrating the data, completing fragmentation, serialization and semantization of the data, extracting or learning entities, attributes and interrelations among the entities from a multi-source database, and forming ontology information expression;
information fusion, integrating the extracted information to eliminate contradiction and ambiguity and produce fragment types including seal bars, terms, indexes, formulas, pictures, tables and annexes;
processing information, namely classifying and storing fragment knowledge generated by the fused information, and constructing a standard knowledge base of the transformer and the breaker after quality evaluation;
specifically, at least 1600 transformer and circuit breaker standard specifications (including national standard, enterprise standard, industry standard, group standard, technical specification, operation instruction book, anti-accident measure, typical design and the like) are processed, including multi-source, heterogeneous data extraction, indexing and production, and integration, so as to complete fragmentation, serialization and semantization of the transformer and circuit breaker, the types of the generated fragments need to include seal bars, terms, indexes, formulas, pictures, tables, appendices and the like, and the processing accuracy is required to be more than 95%; the generated fragment knowledge is classified and stored, a standard knowledge base of the transformer and the breaker is constructed, functions of data warehousing, resource management, resource safety management, version management, system management, data storage, backup and the like are supported, and a shared service capability can be formed and is open to the outside.
The processing standard specification of the transformer and the breaker equipment is as follows: a. chapter bar: extracting and processing the seal strip of each level, supporting the seal strip association nesting, namely hierarchical processing, and the seal strip of the parent level can contain the seal strip of the child level.
b. The terms: and the extraction of the terms is supported, and the extraction result comprises term names, term definitions and the like.
c. Index and index value: and extracting indexes (including character type clauses and numerical types) in the standard and outputting the indexes in a key-value form.
d. The formula: the formula in the standard is extracted, and the formula name and the specific formula are required to be included.
e. Picture: and extracting the non-pure character pictures in the standard, and outputting picture resources which need to include picture names and picture resource files.
f. Table (b): and extracting the table in the standard, wherein the extraction result supports two types of extraction, namely, pictures and excel tables, including table names (if any), table headers (if any), row data, column data and the like.
(2) Constructing a technical standard knowledge base model of the transformer and the breaker equipment:
a. the method comprises the steps of constructing a technical standard knowledge base of the transformer and the breaker equipment, storing data into a digital resource base from the processing and warehousing of resources, managing metadata, digital objects, XML data and the like of the resources, and constructing a technical standard knowledge base of a main network transformer;
b. the system has the functions of data warehousing, resource management, resource security management, version management, system management, data storage, backup and the like, and the abstract sharing service capability is open to the outside.
c. The technical standard knowledge base of the transformer and the breaker equipment needs to comprise a plurality of sub-bases such as a standard file sub-base, a term sub-base, a Zhang Tiao sub-base, an index sub-base, a picture sub-base, a table sub-base and a formula sub-base, wherein each sub-base needs to support data addition, deletion, modification and query, and a front-end graphical page is needed for a user to operate.
In addition, the support service for the standard knowledge graph construction tool model and the standard knowledge graph construction process of the transformer and the breaker equipment comprises the standard knowledge graph construction and increment iteration tool model of the transformer and the breaker equipment, and the whole process technical support, maintenance and training in the aspects of scheme formulation, knowledge modeling, knowledge extraction and review, knowledge disambiguation, graph construction, graph application and the like, and assists in completing the graph construction and updating. The purchased model and service need to run through a business chain from the standard knowledge production of the transformer and the breaker equipment to the knowledge application.
The method for constructing the technical standard knowledge base model of the transformer and the circuit breaker equipment specifically comprises the following steps:
a. the method comprises the steps of constructing a technical standard knowledge base of the transformer and the breaker equipment, starting from the processing and warehousing of resources, storing information data into a digital resource base, managing metadata, digital objects, XML data and the like of the resources, and constructing a technical standard knowledge base of a main network transformer;
b. the system has the functions of data warehousing, resource management, resource safety management, version management, system management, data storage, backup and the like, and the abstract sharing service capability is open to the outside;
c. the technical standard knowledge base of the transformer and the breaker equipment needs to comprise a plurality of sub-bases such as a standard file sub-base, a term sub-base, a Zhang Tiao sub-base, an index sub-base, a picture sub-base, a table sub-base and a formula sub-base, wherein each sub-base needs to support data addition, deletion, modification and query, and a front-end graphical page is needed for a user to operate.
Step 2: constructing an ontology: constructing a mapping layer for representing a mapping relation between a meta-ontology and an ontology in standard data based on the knowledge base obtained in the step 1, wherein the meta-ontology is a common and substantial feature knowledge base extracted from a plurality of ontologies and is used for carrying out abstract expression on the ontology, determining a meta-ontology model and a mapping function between the meta-ontology and the ontology, and correspondingly linking each meta-ontology and each ontology according to the mapping function to construct the mapping layer;
the determining a meta-ontology model and a mapping function between a meta-ontology and an ontology includes: determining a meta-ontology model; determining a first mapping function for characterizing the mapping of meta-ontologies to ontologies according to the meta-ontology model; correspondingly linking each element ontology and each ontology according to the mapping function to construct the mapping layer; constructing a standard knowledge graph based on the mapping layer, wherein the standard knowledge graph comprises the following steps: constructing an ontology model based on the mapping layer; and extracting corresponding entities from the standard text according to the ontology model, and constructing a knowledge graph standard ontology.
The established knowledge graph standard ontology comprises the functions of transformer and circuit breaker equipment standard knowledge extraction and graph establishment: knowledge extraction provides knowledge extraction services aiming at different data sources, and all the knowledge extraction services run periodically in the background in a task mode, so that continuous access of various external data is guaranteed. And the automatic construction from the source database data to the knowledge graph is completed through structured and unstructured data access, and the structured data graph entering capability is provided.
The structured data charting capabilities include:
a) And (3) displaying standard knowledge graphs of the transformer and the breaker equipment: visual map data including entity attributes, relationship query between entities, entity attribute query, and the like; classifying and counting knowledge graph data and visually managing graph data content; supporting upper layer application, providing an interface for inquiring entities, attributes and relationships; the method supports not less than two modeling modes such as lists, visual graphs and the like, and supports graphical entity, relationship and attribute editing;
b) And (3) standard knowledge graph construction and management of the transformer and the breaker equipment: supporting user authority distribution management; visual management map storage is supported; manual intervention or automatic extraction and increase of atlas data (including schema), visual management and traceability of historical atlases are supported;
c) The standard multi-mode knowledge understanding of the transformer and the breaker equipment is as follows: the method supports extraction of knowledge from documents such as PDF, WORD and TXT, and constructs a knowledge graph; aiming at different data forms, text representation information of structural features is used as an analysis object, and text feature calculation and text feature selection are carried out on material contents by utilizing mature technical methods in the fields of machine learning, natural language processing speech recognition, deep learning and the like, combining field problems and actual experience and combining a related database;
d) Other functional requirements are as follows: model training supporting extraction of types such as entities, attributes and the like; supporting the condition of a visual display platform, a training process and result evaluation; providing basic word segmentation and entity recognition capabilities; supporting the corpus labeling capability, enabling a user to customize a label and supporting the labeling of multimode data; the method has a complete knowledge map construction platformization function, and has full stack construction capabilities of knowledge representation, knowledge modeling, knowledge extraction, knowledge fusion, knowledge storage, knowledge calculation and the like; the system has a complete knowledge application platform function and knowledge application capabilities of knowledge retrieval, knowledge question answering and online relation reasoning based on the map.
The constructed knowledge graph standard ontology further comprises a graph storage and query function: the storage, processing and data synchronous updating of structured, semi-structured and other data sources in the process of map construction are supported; supporting relationship management among entities, including adding and deleting edge relationships, setting a plurality of relationship objects and the like; the entity retrieval, entity relation calculation, feature query service and the like of the knowledge graph can be realized; and the knowledge graph content is retrieved and displayed through a complete standard graph query statement.
Referring to FIG. 2, in some embodiments, the ontology-building subsystem is a skeleton layer of the knowledge graph, which defines the basic structure of knowledge, including entity classes, attribute classes, context relationships between entity classes, and ownership relationships between entity attributes. The system adopts a top-down mode to visually construct the knowledge map Schema, supports low-cost custom addition of field attribute information corresponding to various categories, supports presetting of a large number of general knowledge map schemas for system reference, and supports quick generation of the Schema by selecting data from a production source database directly.
The system supports three modes of category creation, manual addition, excel import and synchronization of a data structure of the structured data. The schema created by the three modes is uniformly stored and managed in a schema storage and management module. The system supports the creation of sub-categories under the categories, and the sub-categories automatically inherit the attributes of the parent-level categories, so that the management time of an administrator on the categories with subordinate management is saved.
And the spreadsheet modeling supports manual addition of a category in an interactive spreadsheet operation mode, adds attributes to the category, adds attribute types and constraints, adds relationships, and adds relationship types and constraints.
And mapping modeling, which supports the data structure of synchronous structured data and directly generates the target of knowledge modeling in a fast mapping mode.
The management and display module provides a unified reference, query and modification interface for the schema constructed by the system.
Support for defining complex schemas, including nested representations of attribute values, with attributes defined on edge relationships.
Referring to fig. 3, the knowledge production subsystem inputs various forms of raw data introduced for the data access subsystem and knowledge production targets defined by the ontology construction subsystem, and outputs atlas knowledge.
In some embodiments, the knowledge production subsystem provides the underlying off-line data processing architecture and corresponding support mechanisms. Each type of knowledge production task can be abstracted into two parts: 1) Supporting a unified data processing architecture; 2) A series of policies or algorithms associated with a particular knowledge type. The knowledge production subsystem provides uniform distributed file storage, distributed state storage, distributed result storage, distributed cache, computing scheduling capability, batch processing capability, streaming processing capability and heterogeneous computing capability for the knowledge production strategies or algorithms.
And the map knowledge production module converts the structured data and the unstructured data into knowledge map data and establishes the entity and the relation between the entities. The specific functions include: the method supports the identification of entities, relations and attributes from the free text, and can optimize the accuracy of free text extraction in a mode of manual intervention model; the method supports direct data conversion from a structured data source, aligns with the knowledge map schema mapping, and automatically produces knowledge map data; supporting a user-defined knowledge graph extraction model, including a tuning model, an optimized word list, a definition template and the like; the method supports a machine learning model, machine rules and manual modes to carry out mapping, cleaning, fusion, normalization, edge building and completion of entities, attributes and relationships; the whole process of map knowledge production supports visualization, white box and audit intervention.
The atlas knowledge production module provides overall architecture support by relying on a knowledge production subsystem, and provides the training, executing and predicting capabilities of the algorithm by relying on a model strategy hosting system to complete the serial execution of four big submodules: knowledge extraction, knowledge processing, knowledge fusion and knowledge association.
Processing knowledge, namely performing attribute mapping based on schema on the result of knowledge extraction to ensure that the extracted attribute name conforms to the isonymous attribute defined in the schema; and cleaning the attribute values extracted by the knowledge based on the regular expression so that the extracted attribute values meet the attribute constraint conditions defined in the schema.
Knowledge fusion, the knowledge graph data is often multi-source, the same entities extracted from different sources need to be unified and disambiguated at the instance level, and the homonymous attributes extracted from different sources need to be unified and attribute-preferred at the instance level. The disambiguation strategy of the module mainly realizes that: the method comprises three types of attribute comparison algorithms of text similarity comparison, semantic similarity comparison and various attribute value (such as address, telephone, date and numerical value units) comparison, wherein the upper layer scores the results of the various comparison algorithms through a bank model, an XGboost model and an XGrank model and learns and fits the results.
The module realizes an edge establishing strategy based on rule configuration, and a user can judge whether two entities should be established with an edge relationship based on attributes of types such as character strings, numerical values and the like. And for the established knowledge graph, two modes of rule configuration and knowledge representation learning inference are supported, and potential relations between current knowledge graph entities are completed and mined.
And the knowledge storage subsystem comprises a map storage engine and a text storage engine. And the map storage engine constructs a super-large-scale high-performance distributed map indexing and storage engine. The method supports common Graph models, property graphs and Graph-based Graph query languages similar to Gremlin, provides a Graph native storage engine, supports various storage media/systems, memories or direct SSDs on a storage framework, has distributed storage capacity, meets the storage requirement of mass Graph data, has multiple active instances, is rapid in fault switching, and achieves high availability of services. The text knowledge storage engine integrates the elastic search which is subjected to depth effect and performance optimization, provides a storage and retrieval system for large-scale text data, has expandable system capacity and provides a series of optimizable configurations.
The map storage engine: the graph database BGgraph is a high-performance commercial graph database for Baidu self-research, and is suitable for application scenes of data high correlation and deep analysis. The core of BGgraph is a high-performance graph database engine, which is applied and practiced in a hundred-degree knowledge graph system for years, can support hundreds of millions of entities and millisecond response delay, provides distributed and high availability capability, and meets the requirements of enterprise-level application.
Referring to fig. 4, knowledge base query answering (KB-QA) is one of the most important applications based on a knowledge graph, and refers to a knowledge base oriented to a knowledge graph, which inputs natural language questions, and automatically finds answers from the knowledge graph through query, calculation and reasoning by performing semantic understanding and parsing on the questions, so as to directly meet the needs of users.
In consideration of knowledge storage forms such as SQL databases and tables in enterprises, the system further abstracts and integrates on the technical scheme so as to meet the question and answer scenes of the common structured data knowledge base of the enterprises.
Referring to fig. 5, the knowledge production system of the knowledge base is a core support connecting the base (data management, computation and storage) and the upper application service platform, mainly realizes the functions of intellectual expression, extraction, construction, management and the like of massive multi-source heterogeneous data, and provides knowledge data production support and storage management support for applications such as semantic search, knowledge question answering, computational analysis, inference decision and the like.
The knowledge production system of the knowledge middle platform consists of all core subsystems and supporting and guaranteeing subsystems which run through the whole life cycle of knowledge.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (8)
1. A method for constructing a standard ontology of a knowledge graph is characterized by comprising the following steps:
step 1: and (3) information processing: the transformer and circuit breaker standard specification files in the multi-source database are processed, and the method comprises the following steps:
extracting information, namely extracting multi-source data and heterogeneous data, indexing, producing and integrating the data, completing fragmentation, serialization and semantization of the data, extracting or learning entities, attributes and interrelations among the entities from a multi-source database, and forming ontology information expression;
information fusion, integrating the extracted information to eliminate contradiction and ambiguity and produce fragment types including seal bars, terms, indexes, formulas, pictures, tables and annexes;
processing information, namely classifying and storing fragment knowledge generated by the fused information, and constructing a standard knowledge base of the transformer and the breaker after quality evaluation;
step 2: constructing an ontology: constructing a mapping layer for representing a mapping relation between a meta-ontology and an ontology in standard data based on the knowledge base obtained in the step 1, wherein the meta-ontology is a common and substantial feature knowledge base extracted from a plurality of ontologies and is used for carrying out abstract expression on the ontology, determining a meta-ontology model and a mapping function between the meta-ontology and the ontology, and correspondingly linking each meta-ontology and each ontology according to the mapping function to construct the mapping layer;
the determining a meta-ontology model and a mapping function between a meta-ontology and an ontology includes: determining a meta-ontology model; determining a first mapping function for characterizing the mapping of the meta-ontology to the ontology according to the meta-ontology model; correspondingly linking each element body and each body according to the mapping function to construct the mapping layer; constructing a standard knowledge graph based on the mapping layer, wherein the standard knowledge graph comprises the following steps: constructing an ontology model based on the mapping layer; and extracting corresponding entities from the standard text according to the ontology model, and constructing a knowledge graph standard ontology.
2. The method for constructing the knowledge-graph standard ontology according to claim 1, wherein in the step 1, the generated fragment knowledge is classified and stored, a transformer and a breaker standard knowledge base are constructed, functions of data warehousing, resource management, resource safety management, version management, system management, data storage, backup and the like are supported, and a shared service capability can be formed and is open to the outside.
3. The method for constructing a knowledgegraph standard ontology according to claim 2, wherein the transformer and circuit breaker standard specifications comprise national standards, enterprise standards, industry standards, group standards, technical specifications, work instructions, anti-accident measures, and typical designs.
4. The method for constructing the knowledge-graph standard ontology according to claim 1, wherein in the step 1, the transformer and circuit breaker equipment processing standard specification is as follows:
a. chapter bar: extracting and processing the seal strip of each level, supporting the seal strip to be nested in a correlated manner, namely processing in a hierarchical manner, wherein the seal strip of the parent level can contain the seal strip of the child level;
b. the terms: the method supports the extraction of terms, and the extraction result comprises term names, term definitions and the like;
c. index and index value: extracting indexes in the standard and outputting the indexes in a key-value form;
d. the formula: extracting formulas in the standard, wherein the formulas need to contain formula names and specific formulas;
e. picture: extracting non-pure character pictures in the standard, and outputting picture resources which need to include picture names and picture resource files;
f. table (b): and extracting the table in the standard, wherein the extraction result supports the extraction of two types, namely pictures and excel tables, including table names, table headers, row data, column data and the like.
5. The method for constructing the knowledge-graph standard ontology according to claim 1, wherein in the step 1, a transformer and circuit breaker equipment technical standard knowledge base model is constructed:
a. the method comprises the steps of constructing a technical standard knowledge base of the transformer and the breaker equipment, storing information data into a digital resource base from the processing and warehousing of resources, managing metadata, digital objects, XML data and the like of the resources, and constructing a technical standard knowledge base of a main network transformer;
b. the system has the functions of data warehousing, resource management, resource safety management, version management, system management, data storage, backup and the like, and the abstract sharing service capability is open to the outside;
c. the technical standard knowledge base of the transformer and the breaker equipment needs to comprise a plurality of sub-bases such as a standard file sub-base, a term sub-base, a Zhang Tiao sub-base, an index sub-base, a picture sub-base, a table sub-base and a formula sub-base, wherein each sub-base needs to support data addition, deletion, modification and query, and a front-end graphical page is needed for a user to operate.
6. The method for constructing the knowledge-graph standard ontology according to claim 1, wherein in the step 2, the constructed knowledge-graph standard ontology comprises the functions of extracting the standard knowledge of the transformer and the breaker equipment and constructing the knowledge graph: knowledge extraction provides knowledge extraction services aiming at different data sources, and all the knowledge extraction services are periodically operated in the background in a task mode to ensure the continuous access of various external data; and the automatic construction from the source database data to the knowledge graph is completed through structured and unstructured data access, and the structured data graph entering capability is provided.
7. The method of knowledge-graph standard ontology construction according to claim 6, wherein the structured data mapping capability comprises:
a) And (3) displaying the standard knowledge graph of the transformer and the breaker equipment: visual map data including entity attributes, relationship query between entities, entity attribute query, and the like; classifying and counting knowledge graph data and visually managing graph data content; supporting upper layer application, providing an interface for inquiring entities, attributes and relationships; the method supports not less than two modeling modes such as lists, visual graphs and the like, and supports graphical entity, relationship and attribute editing;
b) And (3) standard knowledge graph construction and management of the transformer and the breaker equipment: supporting user authority distribution management; visual management map storage is supported; manual intervention or automatic extraction and increase of map data, visual management and traceability of historical maps are supported;
c) The standard multi-mode knowledge understanding of the transformer and the breaker equipment is as follows: the method supports extraction of knowledge from documents such as PDF, WORD and TXT, and constructs a knowledge graph; aiming at different data forms, text representation information of structural features is used as an analysis object, and text feature calculation and text feature selection are carried out on material contents by utilizing mature technical methods in the fields of machine learning, natural language processing speech recognition, deep learning and the like, combining field problems and actual experience and combining a related database;
d) Other functional requirements are as follows: model training for extracting types such as entities, attributes and the like is supported; supporting the condition of a visual display platform, a training process and result evaluation; providing basic word segmentation and entity recognition capabilities; supporting corpus labeling capability, enabling a user to customize a label and supporting the labeling of multimode data; the method has a complete knowledge map construction platformization function, and has full stack construction capabilities of knowledge representation, knowledge modeling, knowledge extraction, knowledge fusion, knowledge storage, knowledge calculation and the like; the method has a complete knowledge application platform function, and has knowledge application capabilities of knowledge retrieval, knowledge question answering and online relation reasoning based on the atlas.
8. The method for constructing the knowledge-graph standard ontology according to claim 1, wherein in the step 2, the constructed knowledge-graph standard ontology further comprises a graph storing and querying function: the storage, processing and data synchronous updating of structured, semi-structured and other data sources in the process of map construction are supported; supporting relationship management among entities, including adding and deleting edge relationships, setting a plurality of relationship objects and the like; the entity retrieval, entity relation calculation, feature query service and the like of the knowledge graph can be realized; and the knowledge graph content is retrieved and displayed through a complete standard graph query statement.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211298266.XA CN115757810A (en) | 2022-10-22 | 2022-10-22 | Method for constructing standard ontology of knowledge graph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211298266.XA CN115757810A (en) | 2022-10-22 | 2022-10-22 | Method for constructing standard ontology of knowledge graph |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115757810A true CN115757810A (en) | 2023-03-07 |
Family
ID=85352713
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211298266.XA Pending CN115757810A (en) | 2022-10-22 | 2022-10-22 | Method for constructing standard ontology of knowledge graph |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115757810A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116186232A (en) * | 2023-04-26 | 2023-05-30 | 中国电子技术标准化研究院 | Standard knowledge intelligent question-answering implementation method, device, equipment and medium |
| CN116401410A (en) * | 2023-03-09 | 2023-07-07 | 北京海致星图科技有限公司 | Method, device, storage medium and equipment for accessing map data to multi-scene graph database |
-
2022
- 2022-10-22 CN CN202211298266.XA patent/CN115757810A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116401410A (en) * | 2023-03-09 | 2023-07-07 | 北京海致星图科技有限公司 | Method, device, storage medium and equipment for accessing map data to multi-scene graph database |
| CN116401410B (en) * | 2023-03-09 | 2024-01-26 | 北京海致星图科技有限公司 | Method, device, storage medium and equipment for accessing map data to multi-scene graph database |
| CN116186232A (en) * | 2023-04-26 | 2023-05-30 | 中国电子技术标准化研究院 | Standard knowledge intelligent question-answering implementation method, device, equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yan et al. | KnowIME: a system to construct a knowledge graph for intelligent manufacturing equipment | |
| CN112612902A (en) | Knowledge graph construction method and device for power grid main device | |
| Yang et al. | Research on enterprise risk knowledge graph based on multi-source data fusion | |
| US11928114B2 (en) | Query generation based on a logical data model with one-to-one joins | |
| CN115757810A (en) | Method for constructing standard ontology of knowledge graph | |
| CN111930774B (en) | Automatic construction method and system for electric power knowledge graph body | |
| CN115080765A (en) | Aerospace quality knowledge map construction method, system, medium and equipment | |
| US11928086B2 (en) | Automatic machine learning data modeling in a low-latency data access and analysis system | |
| Hu et al. | EGC: A novel event-oriented graph clustering framework for social media text | |
| Tang et al. | Automatic schema construction of electrical graph data platform based on multi-source relational data models | |
| Gu | Integration and optimization of ancient literature information resources based on big data technology | |
| CN115937881A (en) | Method for automatically identifying content of knowledge graph construction standard form | |
| US11768846B2 (en) | Search guidance | |
| CN115759253A (en) | Power grid operation and maintenance knowledge map construction method and system | |
| Wu et al. | Research of knowledge graph technology and its applications in agricultural information consultation field | |
| US11960484B2 (en) | Identifying joins of tables of a database | |
| Qin et al. | Construction of knowledge graph of multi-source heterogeneous distribution network systems | |
| Meng et al. | Design and Implementation of Knowledge Graph Platform of Power Marketing | |
| Cuddihy et al. | FDC cache: semantics-driven federated caching and querying for big data | |
| Chen et al. | Construction Methods of Knowledge Mapping for Full Service Power Data Semantic Search System | |
| CN113868322B (en) | Semantic structure analysis method, device and equipment, virtualization system and medium | |
| Zhang et al. | The research and design of SQL processing in a data-mining system based on MapReduce | |
| CN115599923A (en) | Power grid standard knowledge extraction method | |
| Fu et al. | Research on knowledge modeling and knowledge push method for space engineering research and development based on ontology | |
| US20240211479A1 (en) | Query Generation Based On A Logical Data Model With One-to-One Joins |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |