Unified data model construction method for steel industry based on big data center

Abstract

The application provides a steel industry unified data model construction method based on a big data center, business is sorted through two dimensions of entity business objects and business processes, industry entity business object and business process sets are sorted out, then clustering and classification are adopted to construct an industry unified theme domain model, an entity business object standard tree and a business process standard tree of the sorted out industry entity business object and business process sets, two inherent relations of combination and derivation existing between the entity business objects are defined in the entity business object tree, relations of the entity business objects due to business activities are defined in the business process, a unified entity business object tree and a business process tree are constructed, and an industry concept graph model is constructed by adopting a knowledge graph technology, and entity definition and relation description in the unified data model construction are clarified.

Description

Technical Field

[0001] This invention relates to the field of data model construction technology, and more specifically, to a method for constructing a unified data model for the steel industry based on a big data center. Background Technology

[0002] Authorized invention patent document CN113407681B discloses a method for constructing a public data model in the energy industry, including the following steps: determining the data subject domain in the public data model; the energy big data center accessing various model data contained in the subject domain; constructing an entity object pool for the energy big data center; constructing a knowledge graph search engine for an energy professional thesaurus; constructing a logical structure model in the public data model, the logical structure model being divided into a core set and a domain set, the core set including a structure model with defined type structures and a basic model, the domain set including a domain model; determining the logical relationship between logical entity i and logical entity j in the domain model according to the knowledge graph search engine; updating the domain model set in real time according to the determination result; and converting the logical structure model into a physical model through a low-code platform.

[0003] However, patent document CN113407681A has the following drawbacks: First, the method for constructing a public data model described in this patent document is only suitable for the energy industry and not for the metallurgical industry. For example, the energy basic information model, business entity ledger model, and social operation data model described in this patent document are not applicable to the metallurgical industry. Second, this patent document only clarifies the dependency relationship and quantitative relationship for the logical entity relationship in the domain model, and cannot reflect the business meaning behind the logical entity relationship. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the purpose of this invention is to provide a method for constructing a unified data model for the steel industry based on a big data center.

[0005] A method for constructing a unified data model for the steel industry based on a big data center, provided by the present invention, includes:

[0006] Step S1: Aggregate and abstract the entity business objects and business process list to obtain the industry subject domain;

[0007] Step S2: Under the industry theme domain, analyze the relationship between entity business objects and business processes, construct a relationship matrix between business processes and entity business objects, and form a corresponding industry concept map model; construct data element standards based on the industry concept map model, and the data element standards are attached to entity business objects to form entity business object models;

[0008] Step S3: In the business process, instantiate the entity business object model based on the real world to generate the instance object table structure;

[0009] Step S4: In the business process, based on business requirements and instance object table structure, identify dimensional attributes and measurement attributes to construct dimensional models and detailed fact models respectively, forming an industry-unified logical data model;

[0010] Step S5: Based on the logical model parser in the big data center, materialize the industry-unified logical data model to generate the actual physical tables of the database.

[0011] Preferably, step S1 includes:

[0012] Step S1.1: Entity business object identification. Based on business requirements, identify the entity business objects that need to be analyzed within the business domain.

[0013] Step S1.2: Entity business object verification and definition. The identified entity business objects are verified according to compliance and completeness. The entity business objects that pass the verification are defined and described to form business object standards.

[0014] Step S1.3: Under the business object standard, identify and define the relationships between entity business objects, and construct the entity business object tree;

[0015] Step S1.4: Business process identification. Based on business requirements and business processes, identify the business processes that need to be analyzed within the business domain.

[0016] Step S1.5: Business process verification and definition. Verify the identified business processes according to compliance and completeness. Define and describe the business processes of entities that pass the verification to form business process standards.

[0017] Step S1.6: Under the business process standard, identify the relationships between entity business processes and construct a business process tree: According to the coarseness of different business processes, clarify the boundaries and relationships between each business process, and construct a business process tree based on the business process relationships; wherein, the business process relationships include composition relationships, and a higher-level business process is composed of one or more lower-level business processes;

[0018] Step S1.7: Define the subject domain. Based on the business process tree and entity business object tree, further cluster and abstract according to the business nature of each business process and entity business object to form an industry subject domain; wherein, the industry subject domain includes a first-level subject domain and a second-level subject domain.

[0019] Preferably, the compliance verification of the entity business object requires the identified entity business object to meet the following three conditions, and the verification is performed by the big data center backend:

[0020] Condition A: The identified entity business object can participate independently in at least one business process;

[0021] Condition B: The identified entity business object should be able to be instantiated, and each instantiated entity business object should have a unique identity identifier in the information system;

[0022] Condition C: The identified entity business object can be relatively independent and has a series of attributes to describe it;

[0023] The aforementioned business object integrity verification refers to ensuring that the identified entity business objects are compatible with business processes, business scenarios, and business management requirements, and that all necessary entity business objects within the business domain are covered.

[0024] The identified entity business objects are classified and clustered, and their boundaries and relationships are clearly defined. A business object tree is constructed based on these relationships. These relationships include composition and derivation. A composition relationship means that a parent node entity business object in the entity business object tree is composed of one or more child node entity business objects. A derivation relationship means that a parent node entity business object in the entity business object tree derives multiple child node entity business objects due to different business scenarios.

[0025] The aforementioned business process compliance verification requires that the identified business processes meet the following two conditions, and is verified by the big data center backend:

[0026] Condition D: The identified business process can be associated with at least two entity business objects;

[0027] Condition E: The identified business process can be associated with at least one metric and one dimension.

[0028] The aforementioned business object integrity verification refers to ensuring that the identified business processes are compatible with business workflows, business scenarios, and business management requirements, thereby covering all necessary business processes within the business domain.

[0029] Preferably, step S2 includes:

[0030] Step S2.1: Analyze the relationship between entity business objects and business processes under the industry subject domain to form a relationship matrix between business processes and entity business objects;

[0031] Step S2.2: Industry concept model construction. Based on the relationship matrix between business processes and entity business objects that has been sorted out, construct an industry concept diagram model between business processes and entity business objects.

[0032] Step S2.3: Establish data element standards based on the industry concept map model;

[0033] Step S2.4: Under the data element standard, based on the object term of the data element, associate the data element with each entity business object tree in the entity business object tree, and define data elements that can uniquely identify the entity business object, forming an entity business object model standard; wherein, the entity business object model standard adopts a multi-branch tree hierarchical model for expression, the tree nodes are divided into root nodes, child nodes and leaf nodes, the root node of the tree is the subject domain, all other child nodes are entity business objects, all leaf nodes are data elements attached to entity business objects, and leaf nodes cannot be further split; the branches of the multi-branch tree represent the relationship between entity business objects, and the relationship attribute is defined as combination or derivation; the branch only connects the adjacent upper and lower level nodes, indicating that the lower level node entity business object is derived from or a sub-component of the upper level node entity business object.

[0034] Preferably, the relationship matrix between the business process and the entity business object is required to satisfy the following two conditions:

[0035] Condition F: The granularity of the business process and the granularity of the objects involved in the entity's business process must be consistent;

[0036] Condition G: Business processes are arranged vertically, and entity business objects are arranged horizontally. Each business process row must be associated with at least two entity business object columns.

[0037] The industry concept diagram model between the business process and the entity business object consists of nodes and links; one node represents one entity business object; a link between two entity business objects represents a business process; each node must have at least one link, that is, each entity business object must participate in at least one business process, and each business process must have a technical expression involving two entity business objects.

[0038] The data element standard includes object terms, feature terms, and representations; the object terms need to reference the entity business objects in the entity business object tree constructed in step S1; when the entity business objects of the parent node and the entity business objects of the child node are in a composition relationship, the object terms in the data element need to reference the entity business objects of the child node, and the entity business objects of the parent node can inherit the entity business object data elements of the child node without redefining new data element standards.

[0039] When the parent node's entity business object and the child node's entity business object have a derived relationship, the child node's entity business object can inherit the parent node's entity business object data elements. During the inheritance process, the parent node's entity business object data element object terms need to be replaced with the child node's entity business object, while the feature terms and representations remain unchanged, and a new data element standard is redefined. The feature terms and representations of the data element standard need to be defined based on the relationship matrix between the business processes and entity business objects identified in the business scenario.

[0040] Preferably, step S3 includes:

[0041] Step S3.1: Identify the entity business objects participating in the business process based on the relationship matrix between business processes and entity business objects;

[0042] Step S3.2: Instantiate the entity business object based on the real world;

[0043] Step S3.3: Under the business process, select the attribute fields to be analyzed for each instance object participating in the business process, and generate the instance object table structure.

[0044] Preferably, the instantiation of entity business objects requires instantiation layer by layer at the node where the entity business objects participating in the business process are located, according to the entity business object tree.

[0045] The instantiation of the entity business object involves renaming the object terms of the data elements associated with the entity business object according to the name of the instantiated instance object, while the characteristic terms and representations remain unchanged. At this time, the data elements are instantiated as the attribute fields of the instance object.

[0046] Preferably, step S4 includes:

[0047] Step S4.1: Under the business process, identify dimensional attributes and construct a dimensional model based on business requirements and instance object table structure;

[0048] Step S4.2: In the business process, identify the measurement attributes based on business requirements and instance object table structure, and construct a detailed fact model by combining the primary key dimension attributes in the dimension model.

[0049] Preferably, the dimensional attributes refer to attributes in the entity business object that are unchanging or whose rate of change is less than a set threshold, providing a description of the business process environment; the dimensional attributes can be split and combined to form new dimensional attributes; the dimensional model needs to determine a primary key for dimensional attributes, and the values ​​of other dimensional attributes in the dimensional model depend on the primary key dimensional attribute; the detailed fact model consists of dimensional attribute foreign keys and the metric attributes of instance objects; the dimensional attribute foreign keys of the detailed fact model are used to connect the primary key dimensional attributes in the dimensional model; the metric attributes of the detailed fact model are used for mathematical statistical analysis and are defined as atomic indicators.

[0050] Preferably, step S5 includes:

[0051] Step S5.1: Materialize the dimensional model based on the logical model parser in the big data center to generate the actual physical tables of the database. Before the materialization process, the logical partitions of the storage database engine and the data storage database must be defined.

[0052] Step S5.2: Materialize the detailed fact model based on the logical model parser in the big data center to generate the actual physical tables of the database. Before the materialization process, the logical partitions of the storage database engine and the data storage database must be defined.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] 1. This invention is suitable for the metallurgical industry. It sorts out business through two dimensions: entity business objects and business processes. In entity business objects, it defines two inherent relationships: combination and derivation between entity business objects. In business processes, it defines the relationships between entity business objects generated by business activities.

[0055] 2. The unified industry logic model for the steel industry constructed by this invention can organize and connect massive amounts of data from various business systems accessing the data center through a unified industry data model, breaking down data silos and analyzing the data from various dimensions to extract useful information. Thus, this invention can quickly provide consistent and reliable indicator analysis dimensions and indicator data sources for data development.

[0056] 3. This invention solves the problem of inconsistent business understanding caused by field name disputes and synonyms in different business systems, which makes it impossible to effectively extract information from data. Attached Figure Description

[0057] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0058] Figure 1 A flowchart illustrating the steps involved in building a unified data model for the steel industry.

[0059] Figure 2 This is a schematic diagram of a partial tree of the device entity business object.

[0060] Figure 3 This is a schematic diagram of a partial view of the material entity business object tree.

[0061] Figure 4 This is a schematic diagram of a partial view of the business process tree.

[0062] Figure 5 This is a schematic diagram of a partial view of the relationship matrix between the hot rolling business process and the entity business object.

[0063] Figure 6 This is a schematic diagram of a partial view of the entity business object model.

[0064] Figure 7 A schematic diagram of a partial diagram instantiated for the entity business object of the finishing mill.

[0065] Figure 8 This is a schematic diagram of a partial view of the unified logic data model for precision rolling. Detailed Implementation

[0066] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0067] This invention first analyzes business operations from two dimensions: entity business objects and business processes. It then identifies a set of industry-specific entity business objects and processes. Next, using clustering and classification methods, it verifies, defines, classifies, and grades these sets, constructing a unified industry-wide subject domain model, a standard tree for entity business objects, and a standard tree for business processes. The entity business object tree defines the inherent relationships of combination and derivation between entity business objects, while the business process defines the relationships between entity business objects arising from business activities. This results in a unified entity business object tree and business process tree. Finally, knowledge graph technology is used to construct an industry concept graph model between business processes and entity business objects, clarifying the entity definitions and relationship descriptions in the construction of the unified data model.

[0068] This invention constructs a data metadata standard attached to entity business objects, enabling standardized descriptions of the attributes of these objects and building a unified entity business object model. Then, by instantiating entity objects, the attributes of each instance are automatically defined during the instantiation process, clarifying the field definitions and descriptions in the construction of the unified data model. This solves the problem of inconsistent business understanding caused by field name discrepancies and synonyms across different business systems, hindering the effective extraction of information from data.

[0069] This invention, based on dimensional modeling technology, constructs a unified industry logical model for the steel industry, comprising a dimensional model and a detailed fact model. The dimensional model uniformly defines the dimensions for data analysis, while the detailed fact model provides reliable and consistent measurement information. The unified industry logical model for the steel industry constructed using this invention can quickly provide consistent and reliable indicator analysis dimensions and indicator data sources for data development.

[0070] Specifically, this invention provides a method for constructing a unified data model for the steel industry based on a big data center, the steps of which are as follows: Figure 1 As shown, it includes:

[0071] Step S1: Obtain a list of entity business objects and business processes through business analysis, further aggregate and abstract the list of entity business objects and business processes, and combine industry knowledge to obtain industry subject domains.

[0072] Step S1.1: Entity business object identification. Based on business requirements, identify the entity business objects that need to be analyzed within the business domain.

[0073] Step S1.2: Entity business object verification and definition. The identified entity business objects are verified according to compliance and completeness. The verified entity business objects are defined and described to form business object standards.

[0074] The compliance verification of the aforementioned entity business objects requires that the identified entity business objects meet the following three conditions, and the big data center backend performs algorithm verification:

[0075] Condition A: The identified entity business object can participate independently in at least one business process;

[0076] Condition B: The identified entity business object should be able to be instantiated, and each instantiated entity business object should have a unique identifier in the information system.

[0077] Condition C: The identified entity business object can be relatively independent and has a series of attributes to describe it.

[0078] The aforementioned business object integrity verification refers to ensuring that the identified entity business objects are compatible with business processes, business scenarios, and business management requirements, and that all necessary entity business objects within the business domain are covered.

[0079] Step S1.3: Identify and define the relationships between entity business objects, construct an entity business object tree, classify and cluster the identified entity business objects, clarify and define the boundaries and relationships between entity business objects, and construct a business object tree based on the relationships between entity business objects.

[0080] The entity business object relationships include composition relationships and derivation relationships.

[0081] The aforementioned composition relationship refers to the fact that a parent node entity business object in the entity business object tree can be composed of multiple child node entity business objects.

[0082] The aforementioned derivation relationship refers to the fact that a parent node entity business in the entity business object tree can derive multiple child node entity business objects due to different business scenarios.

[0083] Example: A device entity business object tree was constructed, a part of which is shown below. Figure 2As shown, the hot rolling mill entity business object is a node, and it is composed of sub-node entity business objects such as heating furnace, roughing mill, finishing mill, tilting machine, and coiling machine.

[0084] Example: A material entity business object tree was constructed, a part of which is shown below. Figure 3 As shown, the material entity business object is a node, and it has a derived relationship with the sub-node entity business objects such as molten iron, molten steel, billet, hot rolled coil, and cold rolled coil.

[0085] Step S1.4: Business process identification. Based on business requirements and business processes, identify the business processes that need to be analyzed within the business domain.

[0086] Step S1.5: Business process verification and definition. Verify the identified business processes according to compliance and completeness. Define and describe the verified entity business processes to form business process standards.

[0087] The aforementioned business process compliance verification requires that the identified business processes meet the following two conditions, and the big data center backend performs algorithm verification:

[0088] Condition D: The identified business process can be associated with at least two entity business objects;

[0089] Condition E: The identified business process can be associated with at least one metric and one dimension.

[0090] The aforementioned business object integrity verification refers to ensuring that the identified business processes are compatible with business workflows, business scenarios, and business management requirements, thereby covering all necessary business processes within the business domain.

[0091] Step S1.6: Identify entity business process relationships and construct a business process tree: Based on the coarse-grained nature of different business processes, clarify the boundaries and relationships between each business process, and construct a business process tree based on the business process relationships.

[0092] The business process relationships mentioned include composition relationships, where a higher-level business process can be composed of multiple lower-level business processes.

[0093] Example: A business process tree for the manufacturing domain was constructed, a part of which is shown below. Figure 4 As shown, the hot rolling process is a node, which is composed of sub-node processes such as heating, roughing, finishing, laminar cooling, and coiling.

[0094] Step S1.7: Define the subject domain. Based on the business process tree and entity business object tree, further cluster and abstract according to the business nature of each business process and entity business object to form the industry subject domain.

[0095] The industry subject domains include primary subject domains and secondary subject domains.

[0096] Example: The entity business objects such as hot rolling mill, cold rolling mill, steelmaking unit, materials, raw materials, and production plan constructed in steps S1.3 and S1.4, as well as the business processes such as hot rolling, cold rolling, steelmaking, inspection and testing, and plan issuance, are clustered and abstracted to form the manufacturing primary subject domain.

[0097] Step S2: Under the subject domain, analyze the relationship between entity business objects and business processes, construct a relationship matrix between entity business processes and entity business objects, and form an industry concept model; construct data element standards based on the relationship matrix between business processes and entity business objects, and the data element standards are attached to entity business objects to form an entity business object model.

[0098] Step S2.1: Industry concept model construction, analyze the relationship between entity business objects and business processes under the subject domain, and form a relationship matrix between business processes and entity business objects.

[0099] The relationship matrix between the business process and the entity business object must satisfy the following two conditions:

[0100] Condition F: The granularity of the business process and the granularity of the objects involved in the entity's business process must be consistent.

[0101] Condition G: Business processes are arranged vertically, and entity business objects are arranged horizontally. Each business process row must be associated with at least two entity business object columns.

[0102] Example: Hot rolling business process—entity business object relationship matrix as follows Figure 5 As shown, the entity business objects of the hot rolling mill and the hot rolling business process have the same granularity and are marked as participating in the matrix. However, if the entity business objects of the hot rolling mill and the heating business process have different granularities, no relationship can be generated to form a matrix. The heating furnace and the heating business process have the same granularity and can generate a relationship to form a matrix.

[0103] Business Process - Partial Table of Entity Business Object Relationship Matrix

[0104]

[0105] Step S2.2: Industry concept model construction. Based on the relationship matrix between the business processes and entity business objects that has been sorted out, construct an industry concept graph model between the business processes and entity business objects.

[0106] The industry concept graph model between the business processes and entity business objects consists of nodes and links.

[0107] In the industry concept graph model between the business process and the entity business object, a node represents an entity business object.

[0108] In the industry concept graph model between the business process and the entity business object, the link between two entity business objects represents a business process.

[0109] In the industry concept graph model between the business process and the entity business object, each node must have at least one link, that is, each entity business object must participate in at least one business process, and each business process must have a technical expression involving two entity business objects.

[0110] Example: The graph model constructed based on the partial table of the business process-entity business object relationship matrix is ​​shown below:

[0111] Step S2.3: Construct data element standards based on the relationship matrix between business processes and entity business objects.

[0112] The data element standard mentioned above refers to the data element model of GBXXXX and consists of object words, feature words, and representations.

[0113] The object term mentioned above needs to reference the entity business object in the entity business object tree constructed in step S1.3.

[0114] When the entity business object of the parent node and the entity business object of the child node are in a compositional relationship, the object term in the data element needs to reference the entity business object of the child node. The entity business object of the parent node can inherit the entity business object data element of the child node without redefining a new data element standard.

[0115] When the entity business object of the parent node and the entity business object of the child node are in a derived relationship, the entity business object of the child node can inherit the entity business object data elements of the parent node. During the inheritance process, the data element object words of the parent node entity business object need to be replaced with the data element object words of the child node entity business object, while the feature words and representations remain unchanged, and a new data element standard is redefined.

[0116] The feature words and representations of the data element standard need to be defined based on the relationship matrix between the business processes and entity business objects identified in the business scenario.

[0117] Step S2.4: Based on the object terms of the data elements, associate the data elements with each entity business object tree in the entity business object tree, and define data elements that can uniquely identify the entity business object to form the entity business object model standard.

[0118] The entity business object model standard described above uses a multi-way tree hierarchical model for representation. The tree nodes are divided into root nodes, child nodes, and leaf nodes, as shown below. Figure 6As shown, the root node of the tree is the subject domain, all other child nodes are entity business objects, and all leaf nodes are data elements attached to entity business objects. Leaf nodes cannot be further divided. The branches of the multi-way tree represent the relationships between entity business objects, and the relationship attributes can be defined as composition or derivation. A branch only connects adjacent nodes at the upper and lower levels, indicating that the entity business object of the lower-level node is derived from or a sub-component of the entity business object of the upper-level node.

[0119] The present invention uses a hierarchical structure of multi-branch trees to express the entity business object model, which can decompose and combine each entity business object at each level, and perform individual analysis and mathematical modification of each node entity business object according to requirements.

[0120] Step S3: In the business process, instantiate the entity business object model based on the real world to generate the instance object table structure.

[0121] Step S3.1: Identify the entity business objects participating in the business process based on the relationship matrix between the business process and the entity business object.

[0122] Step S3.2: Instantiate the entity business object based on the real world.

[0123] The instantiation of entity business objects requires instantiation layer by layer at the node where the entity business object participating in the business process is located, according to the entity business object tree.

[0124] The instantiation of the entity business object involves renaming the object terms of the data elements associated with the entity business object according to the name of the instantiated instance object, while the characteristic terms and representations remain unchanged. At this time, the data elements are instantiated as the attribute fields of the instance object.

[0125] Example: Figure 7 This is a partial diagram showing the instantiation of a finishing mill entity business object in a specific application scenario. The finishing mill has multiple stands, each instantiated from the same stand entity business object, inheriting all data elements from the stand entity business object to generate the instance object's attribute information. The work rolls of each stand are instantiated as upper and lower work roll instance objects, inheriting all data elements from the work roll entity business object to generate the instance object's attribute information.

[0126] Step S3.3: Under the business process, select the attribute fields to be analyzed for each instance object participating in the business process, and generate the instance object table structure.

[0127] Step S4: In the business process, based on business requirements and instance object table structure, identify dimensional attributes and measurement attributes, and construct dimensional models and detailed fact models respectively to form an industry-unified logical data model.

[0128] Step S4.1: In the business process, identify dimensional attributes and construct a dimensional model based on business requirements and instance object table structure.

[0129] The aforementioned dimensional attributes refer to the attributes in an entity business object that are unchanging or slowly changing, providing a description of the business process environment.

[0130] The dimensional attributes mentioned above can also be split and combined to form new dimensional attributes.

[0131] The dimensional model requires a primary key for a dimensional attribute, and the values ​​of other dimensional attributes in the dimensional model depend on the primary key dimensional attribute.

[0132] Step S4.2: In the business process, identify the measurement attributes based on business requirements and instance object table structure, and construct a detailed fact model by combining the primary key dimension attributes in the dimension model.

[0133] The detailed fact model consists of dimensional attribute foreign keys and instance object metric attributes.

[0134] The foreign key of the dimension attribute in the detailed fact model can be used to link the primary key dimension attribute in the dimension model.

[0135] The metric attributes of the detailed fact model described above can be used for mathematical statistical analysis and are defined as atomic indicators.

[0136] Example: Figure 8 As shown, a partial diagram of the unified logical model of finishing rolling is constructed under the finishing rolling business process. The fact model of this model is the finishing rolling fact model, and the dimension model consists of four dimensions: date dimension, unit dimension, steel tapping mark dimension, and defect dimension.

[0137] Step S5: Based on the logical model parser in the big data center, materialize the unified logical data model to generate the actual physical tables of the database.

[0138] Step S5.1: First, materialize the dimensional model based on the logical model parser in the big data center to generate the actual physical tables of the database. Before the materialization process, the logical partitions of the storage database engine and the data storage database must be defined.

[0139] Step S5.2: Next, based on the logical model parser in the big data center, the detailed fact model is materialized to generate the actual physical tables of the database. Before the materialization process, the logical partitions of the storage database engine and the data storage database must be defined.

[0140] Those skilled in the art will understand that, in addition to implementing the system, apparatus, and their modules provided by this invention in purely computer-readable program code, the same program can be implemented in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system, apparatus, and their modules provided by this invention can be considered a hardware component, and the modules included therein for implementing various programs can also be considered structures within the hardware component; alternatively, modules for implementing various functions can be considered both software programs implementing the method and structures within the hardware component.

[0141] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A method for constructing a unified data model for the steel industry based on a big data center, characterized in that, include: Step S1: Aggregate and abstract the entity business objects and business process list to obtain the industry subject domain; Step S2: Under the industry theme domain, analyze the relationship between entity business objects and business processes, construct a relationship matrix between business processes and entity business objects, and form a corresponding industry concept map model; construct data element standards based on the industry concept map model, and the data element standards are attached to entity business objects to form entity business object models; Step S3: In the business process, instantiate the entity business object model based on the real world to generate the instance object table structure; Step S4: In the business process, based on business requirements and instance object table structure, identify dimensional attributes and measurement attributes to construct dimensional models and detailed fact models respectively, forming an industry-unified logical data model; Step S5: Based on the logical model parser in the big data center, materialize the industry-unified logical data model to generate the actual physical tables of the database.

2. The method for constructing a unified data model for the steel industry based on a big data center according to claim 1, characterized in that, Step S1 includes: Step S1.1: Entity business object identification. Based on business requirements, identify the entity business objects that need to be analyzed within the business domain. Step S1.2: Entity business object verification and definition. The identified entity business objects are verified according to compliance and completeness. The entity business objects that pass the verification are defined and described to form business object standards. Step S1.3: Under the business object standard, identify and define the relationships between entity business objects, and construct the entity business object tree; Step S1.4: Business process identification. Based on business requirements and business processes, identify the business processes that need to be analyzed within the business domain. Step S1.5: Business process verification and definition. Verify the identified business processes according to compliance and completeness. Define and describe the business processes of entities that pass the verification to form business process standards. Step S1.6: Under the business process standard, identify the relationships between entity business processes and construct a business process tree: According to the coarseness of different business processes, clarify the boundaries and relationships between each business process, and construct a business process tree based on the business process relationships; wherein, the business process relationships include composition relationships, and a higher-level business process is composed of one or more lower-level business processes; Step S1.7: Define the subject domain. Based on the business process tree and entity business object tree, further cluster and abstract according to the business nature of each business process and entity business object to form an industry subject domain; wherein, the industry subject domain includes a first-level subject domain and a second-level subject domain.

3. The method for constructing a unified data model for the steel industry based on a big data center according to claim 2, characterized in that, The compliance verification of the aforementioned entity business objects requires that the identified entity business objects meet the following three conditions, and the verification is performed by the big data center backend: Condition A: The identified entity business object can participate independently in at least one business process; Condition B: The identified entity business object should be able to be instantiated, and each instantiated entity business object should have a unique identity identifier in the information system; Condition C: The identified entity business object can be relatively independent and has a series of attributes to describe it; The aforementioned business object integrity verification refers to ensuring that the identified entity business objects are compatible with business processes, business scenarios, and business management requirements, and that all necessary entity business objects within the business domain are covered. The identified entity business objects are classified and clustered, and their boundaries and relationships are clearly defined. A business object tree is constructed based on these relationships. These relationships include composition and derivation. A composition relationship means that a parent node entity business object in the entity business object tree is composed of one or more child node entity business objects. A derivation relationship means that a parent node entity business object in the entity business object tree derives multiple child node entity business objects due to different business scenarios. The aforementioned business process compliance verification requires that the identified business processes meet the following two conditions, and is verified by the big data center backend: Condition D: The identified business process can be associated with at least two entity business objects; Condition E: The identified business process can be associated with at least one metric and one dimension. The aforementioned business object integrity verification refers to ensuring that the identified business processes are compatible with business workflows, business scenarios, and business management requirements, thereby covering all necessary business processes within the business domain.

4. The method for constructing a unified data model for the steel industry based on a big data center according to claim 3, characterized in that, Step S2 includes: Step S2.1: Analyze the relationship between entity business objects and business processes under the industry subject domain to form a relationship matrix between business processes and entity business objects; Step S2.2: Industry concept model construction. Based on the relationship matrix between business processes and entity business objects that has been sorted out, construct an industry concept diagram model between business processes and entity business objects. Step S2.3: Establish data element standards based on the industry concept map model; Step S2.4: Under the data element standard, based on the object term of the data element, associate the data element with each entity business object tree in the entity business object tree, and define data elements that can uniquely identify the entity business object, forming an entity business object model standard; wherein, the entity business object model standard adopts a multi-branch tree hierarchical model for expression, the tree nodes are divided into root nodes, child nodes and leaf nodes, the root node of the tree is the subject domain, all other child nodes are entity business objects, all leaf nodes are data elements attached to entity business objects, and leaf nodes cannot be further split; the branches of the multi-branch tree represent the relationship between entity business objects, and the relationship attribute is defined as combination or derivation; the branch only connects the adjacent upper and lower level nodes, indicating that the lower level node entity business object is derived from or a sub-component of the upper level node entity business object.

5. The method for constructing a unified data model for the steel industry based on a big data center according to claim 4, characterized in that, The relationship matrix between the business process and the entity business object must satisfy the following two conditions: Condition F: The granularity of the business process and the granularity of the objects involved in the entity's business process must be consistent; Condition G: Business processes are arranged vertically, and entity business objects are arranged horizontally. Each business process row must be associated with at least two entity business object columns. The industry concept diagram model between the business process and the entity business object consists of nodes and links; one node represents one entity business object; a link between two entity business objects represents a business process; each node must have at least one link, that is, each entity business object must participate in at least one business process, and each business process must have a technical expression involving two entity business objects. The data element standard includes object terms, feature terms, and representations; the object terms need to reference the entity business objects in the entity business object tree constructed in step S1; when the entity business objects of the parent node and the entity business objects of the child node are in a composition relationship, the object terms in the data element need to reference the entity business objects of the child node, and the entity business objects of the parent node can inherit the entity business object data elements of the child node without redefining new data element standards. When the entity business object of the parent node and the entity business object of the child node have a derived relationship, the entity business object of the child node can inherit the entity business object data elements of the parent node. During the inheritance process, the data element object words of the parent node entity business object need to be replaced with the entity business object of the child node, the feature words and the representation remain unchanged, and a new data element standard is redefined. The feature words and representations of the data element standard need to be defined based on the relationship matrix between the business process and the entity business object according to the business scenario.

6. The method for constructing a unified data model for the steel industry based on a big data center according to claim 1, characterized in that, Step S3 includes: Step S3.1: Identify the entity business objects participating in the business process based on the relationship matrix between business processes and entity business objects; Step S3.2: Instantiate the entity business object based on the real world; Step S3.3: Under the business process, select the attribute fields to be analyzed for each instance object participating in the business process, and generate the instance object table structure.

7. The method for constructing a unified data model for the steel industry based on a big data center according to claim 6, characterized in that, The instantiation of entity business objects requires instantiation layer by layer at the node where the entity business objects participating in the business process are located, according to the entity business object tree. The instantiation of the entity business object involves renaming the object terms of the data elements associated with the entity business object according to the name of the instantiated instance object, while the characteristic terms and representations remain unchanged. At this time, the data elements are instantiated as the attribute fields of the instance object.

8. The method for constructing a unified data model for the steel industry based on a big data center according to claim 1, characterized in that, Step S4 includes: Step S4.1: Under the business process, identify dimensional attributes and construct a dimensional model based on business requirements and instance object table structure; Step S4.2: In the business process, identify the measurement attributes based on business requirements and instance object table structure, and construct a detailed fact model by combining the primary key dimension attributes in the dimension model.

9. The method for constructing a unified data model for the steel industry based on a big data center according to claim 8, characterized in that, The aforementioned dimensional attributes refer to attributes in entity business objects that remain unchanged or whose rate of change is less than a set threshold, providing a description of the business process environment; the aforementioned dimensional attributes can be split and combined to form new dimensional attributes; the aforementioned dimensional model needs to determine a primary key for dimensional attributes, and the values ​​of other dimensional attributes in the dimensional model depend on the primary key dimensional attribute; the aforementioned detailed fact model consists of dimensional attribute foreign keys and the metric attributes of instance objects; the dimensional attribute foreign keys of the aforementioned detailed fact model are used to link the primary key dimensional attributes in the dimensional model; the metric attributes of the aforementioned detailed fact model are used for mathematical statistical analysis and are defined as atomic indicators.

10. The method for constructing a unified data model for the steel industry based on a big data center according to claim 8, characterized in that, Step S5 includes: Step S5.1: Materialize the dimensional model based on the logical model parser in the big data center to generate the actual physical tables of the database. Before the materialization process, the logical partitions of the storage database engine and the data storage database must be defined. Step S5.2: Materialize the detailed fact model based on the logical model parser in the big data center to generate the actual physical tables of the database. Before the materialization process, the logical partitions of the storage database engine and the data storage database must be defined.

Images