Method and system for constructing graph database based on multi-dimensional and multi-state digital architecture

Abstract

The embodiment of the invention provides a method and system for constructing a graph database based on a multi-dimensional and multi-state digital architecture, and belongs to the technical field of graph databases. The method comprises the following steps: acquiring multi-dimensional and multi-state digital architecture data, and preprocessing the acquired data; after the preprocessing is completed, performing entity identification and merging on the multi-dimensional and multi-state digital architecture data; when entity identification and combination are carried out on the multi-dimensional and multi-state digital architecture data, attribute fusion is carried out on identified entities; based on entity recognition and attribute fusion, fusing the relationships between the entities; and constructing a graph database based on the entities after attribute fusion and relationship fusion, and optimizing and regularly updating the constructed database. According to the method, a large amount of data of the architecture assets can be collected, relation fusion is carried out, and then the visual node graph is constructed, so that the relevance between the architecture assets can be visually displayed.

Description

Technical Field

[0001] This invention relates to the field of graph database technology, and more specifically to a method and system for constructing graph databases based on a multidimensional polymorphic digital architecture. Background Technology

[0002] In existing technologies, architectural assets are primarily stored and retrieved in the form of electronic documents, and are scattered across different departments, units, and individuals. These assets not only have diverse data sources and formats, but also lack unified integration standards and specifications. This not only significantly increases the difficulty of using architectural assets, but also severely restricts the quality and efficiency of architectural design.

[0003] Currently, a large number of architectural assets exist. Traditional relational databases are often inefficient in processing semi-structured assets, lacking the ability to deeply integrate architectural assets of different types and sources. They fail to establish organic connections between these assets, making it difficult to effectively conduct data analysis across business domains and asset categories. This hinders the support of enterprises for global analysis and decision-making in complex business scenarios, significantly limiting the value release of architectural assets. Therefore, a graph database construction method based on a multidimensional and polymorphic digital architecture is needed to collect data from a large number of architectural assets, integrate relationships, and then construct an intuitive node graph to visually display the relationships between various architectural assets. Summary of the Invention

[0004] The purpose of this invention is to provide a method for constructing a graph database based on a multidimensional polymorphic digital architecture. This method can collect a large amount of data on architectural assets, perform relationship fusion, and then construct an intuitive node graph to visually display the relationships between various architectural assets.

[0005] To achieve the above objectives, embodiments of the present invention provide a method for constructing a graph database based on a multidimensional polymorphic digital architecture, the method comprising: Acquire multidimensional and polymorphic digital architecture data and preprocess the acquired data; After preprocessing, entity recognition and merging are performed on the multidimensional polymorphic digital architecture data; While performing entity identification and merging on the multidimensional and polymorphic digital architecture data, attribute fusion is performed on the identified entities; The relationships between the entities are fused based on entity recognition and attribute fusion; A graph database is constructed based on entities obtained through attribute fusion and relationship fusion, and the constructed database is optimized and updated regularly.

[0006] Optionally, multidimensional and polymorphic digital architecture data is acquired, and the acquired data is preprocessed, including: Acquire the multidimensional polymorphic digital architecture data and perform data cleaning on the multidimensional polymorphic digital architecture data; After cleaning, the relational tables, XML files, and JSON objects in the cleaned multidimensional polymorphic digital architecture data are converted into node and edge structures suitable for graph databases; The numerical data in the multidimensional polymorphic digital architecture data are normalized.

[0007] Optionally, after preprocessing, entity identification and merging are performed on the multidimensional polymorphic digital architecture data, including: Obtain the entities to be identified and their corresponding attributes from the multidimensional polymorphic digital architecture data; Based on the entities to be identified and their corresponding attributes, construct a node graph; The similarity between two nodes is calculated based on the attributes in the node graph. Obtain the neighbor set corresponding to the node, and calculate the semantic path similarity between the two nodes using formula (1): Formula (1), in, Represents a node and nodes The similarity of semantic paths between them Representing a path The weight, Indicates the connection node and nodes The set of semantic paths Represents a node The degree, Represents a node The degree; Based on the obtained similarity between nodes and the similarity of semantic paths between nodes, the matching similarity between nodes is calculated using formula (2): Formula (2), in, Represents a node and nodes Matching similarity, Indicates the balancing weights. Represents a node and nodes The similarity between them; Based on the obtained matching similarity between nodes, two nodes with a matching similarity greater than a preset threshold are selected and merged to obtain a new node; After the merge is complete, the edges that were connected to the original nodes are connected to the new nodes.

[0008] Optionally, calculating the similarity between two nodes based on attributes in the node graph includes: Get the two nodes to be matched, and get the attributes of the corresponding nodes; The similarity between two nodes is calculated using formula (3) based on the obtained attributes: Formula (3), in, Indicates the total number of attributes. Represents attributes The weight, This indicates an indicator function, which has attributes in both nodes. The value is 1 when it is not empty, and 0 in all other cases. Indicates a single attribute The similarity between the two nodes.

[0009] Optionally, based on entity recognition and attribute fusion, the relationships between the entities are fused, including: Obtain the identified entities and their corresponding attributes, as well as the various relationships between the entities and their corresponding attributes; The entities and the various relationships between them are fed into a relationship ranking model to rank the relationship levels between the entities. After sorting, determine whether there are any conflicting relationships between the entities in the sorting; When there are conflicting relationships between entities in the sorting process, discard the relationships between entities that are ranked lower. The relationships between the sorted entities are weighted, then merged, and the relationships between the entities that are ranked first are highlighted.

[0010] Optionally, the entities and various relationships between them are fed into a relationship ranking model to rank the relationship levels between entities, including: The relationship ranking model calculates the ranking score based on the relationships between entities using formula (4): Formula (4), in, Indicates the sorting score. Positional parameters representing the relationships between entities in the sorting process. This indicates the number of relationships between entities in the sorting. Indicates the first The relevance of relationships between entities; The relationships between entities are sorted in an ideal order, and their corresponding parameters are imported into the formula (4) to obtain the ideal sorting score. The normalized cumulative gain of loss is calculated using formula (5) based on the obtained ideal ranking score: Formula (5), in, This represents the normalized cumulative gain. Indicates the score of the ideal sort; The goal is to maximize the cumulative gain of the normalized loss in order to obtain the trained relational ranking model. The entities that need to be sorted and the various relationships between them are fed into the trained relationship ranking model to sort the relationship levels between the entities.

[0011] Optionally, the relevance of relationships between entities can be obtained, including: The confidence level, reliability, timeliness, and number of pieces of evidence for obtaining the source of the relationship between entities; The relevance of relationships between entities is classified into levels such as unrelated, somewhat related, related, highly related, very related, and perfectly related, thus obtaining a level matrix; An initial association matrix is ​​obtained based on the confidence, reliability, timeliness, and membership degree of the source of the relationships between entities and the number of evidence and the level of the classification. The obtained initial correlation matrices are aggregated to obtain a fuzzy relation matrix; Weights are assigned to the confidence level, reliability, timeliness, and number of evidence of the source of the relationship between entities to obtain a weight matrix; The weight matrix is ​​multiplied by the fuzzy relation matrix to obtain the comprehensive correlation matrix; Based on the obtained comprehensive association matrix, it is matched with the level matrix to obtain the level of relevance between the corresponding entities, thereby obtaining the corresponding relevance.

[0012] Optionally, the multidimensional polymorphic digital architecture data includes structured, semi-structured, and unstructured data. Structured data is stored in business process systems, application servers, and database management system platforms, while semi-structured and unstructured data are scattered in various documents, such as project requirements specifications, feasibility study reports, and preliminary designs, which contain a wealth of data information.

[0013] On the other hand, the present invention can also provide a system for constructing a graph database based on a multidimensional polymorphic digital architecture, the system comprising: The data acquisition module is used to acquire multidimensional and polymorphic digital architecture data; The module is used to construct a graph database based on the acquired multidimensional polymorphic digital architecture data and execute the method described above for constructing a graph database based on the multidimensional polymorphic digital architecture.

[0014] Through the above technical solution, the method and system for constructing a graph database based on a multidimensional polymorphic digital architecture provided by this invention acquires multidimensional polymorphic digital architecture data and then preprocesses the acquired data. After preprocessing, entity identification and merging can be performed on the multidimensional polymorphic digital architecture data, thereby merging some duplicate entities. Simultaneously with entity identification and merging, attribute fusion can be performed on the identified entities. This attribute fusion can be based on the attributes corresponding to the identified entities or other methods. After entity identification and attribute fusion, the relationships between entities can be fused based on the entity identification and attribute fusion. After fusion, a graph database can be constructed based on the entities resulting from attribute fusion and relationship fusion. After the database is constructed, it can be optimized and updated periodically as needed. This method can collect data on a large number of architectural assets, perform relationship fusion, and then construct an intuitive node graph to visually display the relationships between various architectural assets.

[0015] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0016] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of a method for constructing a graph database based on a multidimensional polymorphic digital architecture according to an embodiment of the present invention; Figure 2 This is a flowchart of the preprocessing steps for a method of constructing a graph database based on a multidimensional polymorphic digital architecture according to an embodiment of the present invention. Figure 3 This is a flowchart of entity recognition and merging in a method for constructing a graph database based on a multidimensional polymorphic digital architecture according to an embodiment of the present invention; Figure 4 This is a flowchart illustrating the relationship fusion process of a method for constructing a graph database based on a multidimensional polymorphic digital architecture according to an embodiment of the present invention. Figure 5 This is a flowchart illustrating the hierarchical sorting of a method for constructing a graph database based on a multidimensional polymorphic digital architecture according to an embodiment of the present invention. Figure 6 This is a flowchart illustrating the process of obtaining relevance in a method for constructing a graph database based on a multidimensional polymorphic digital architecture according to an embodiment of the present invention. Detailed Implementation

[0017] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0018] In the embodiments of this application, certain software, components, models and other existing solutions in the industry may be mentioned. These should be regarded as exemplary and are only intended to illustrate the feasibility of implementing the technical solution of this application. However, they do not mean that the applicant has used or necessarily used the solution.

[0019] Figure 1 This is a flowchart of a method for constructing a graph database based on a multidimensional polymorphic digital architecture according to an embodiment of the present invention. In this invention, the process of constructing the graph database may include: In step S1, multidimensional and polymorphic digital architecture data is acquired, and the acquired data is preprocessed.

[0020] In step S2, after preprocessing, entity identification and merging are performed on the multidimensional polymorphic digital architecture data.

[0021] In step S3, while performing entity identification and merging on the multidimensional polymorphic digital architecture data, attribute fusion is performed on the identified entities.

[0022] In step S4, the relationships between entities are fused based on entity recognition and attribute fusion.

[0023] In step S5, a graph database is constructed based on the entities after attribute fusion and relationship fusion, and the constructed database is optimized and updated regularly.

[0024] In this invention, when constructing a graph database, multidimensional and polymorphic digital architecture data can be acquired, and then preprocessed. After preprocessing, entity identification and merging can be performed on the multidimensional and polymorphic digital architecture data to merge some duplicate entities. Simultaneously with entity identification and merging, attribute fusion can be performed on the identified entities. This attribute fusion can be based on the corresponding attributes of the identified entities or other methods. After entity identification and attribute fusion, the relationships between entities can be fused based on this entity identification and attribute fusion. After fusion, a graph database can be constructed based on the entities resulting from attribute fusion and relationship fusion. After the database is constructed, it can be optimized and updated periodically as needed. This method can collect data on a large number of architectural assets, perform relationship fusion, and then construct an intuitive node graph to visually display the relationships between various architectural assets.

[0025] In one embodiment of the present invention, such as Figure 2 As shown, the preprocessing flow may include: In step S6, multidimensional polymorphic digital architecture data is acquired and data cleaning is performed on the multidimensional polymorphic digital architecture data.

[0026] In step S7, after the cleaning is completed, the relational tables, XML files, and JSON objects in the cleaned multidimensional polymorphic digital architecture data are converted into node and edge structures suitable for graph databases.

[0027] In step S8, the numerical data in the multidimensional polymorphic digital architecture data is normalized.

[0028] In this invention, when preprocessing the acquired data, the multidimensional polymorphic digital architecture data can be obtained first, and then data cleaning can be performed on the multidimensional polymorphic digital architecture data. After cleaning, the relational tables, XML files, and JSON objects in the cleaned multidimensional polymorphic digital architecture data can be converted into node and edge structures suitable for graph databases, so as to facilitate the subsequent construction of graph databases. Then, the numerical data in the multidimensional polymorphic digital architecture data can be normalized, thereby completing the preprocessing of the data.

[0029] In one embodiment of the present invention, such as Figure 3 As shown, the entity recognition and merging process may include: In step S9, the entities to be identified and their corresponding attributes are obtained from the multidimensional polymorphic digital architecture data.

[0030] In step S10, a node graph is constructed based on the entity to be identified and its corresponding attributes.

[0031] In step S11, the similarity between two nodes is calculated based on the attributes in the node graph.

[0032] In step S12, the neighbor set corresponding to the node is obtained, and the semantic path similarity between the two nodes is calculated using formula (1): Formula (1), in, Represents a node and nodes The similarity of semantic paths between them Representing a path The weight, Indicates the connection node and nodes The set of semantic paths Represents a node The degree, Represents a node The degree.

[0033] In step S13, based on the obtained similarity between nodes and the similarity of semantic paths between nodes, the matching similarity between nodes is calculated using formula (2): Formula (2), in, Represents a node and nodes Matching similarity, Indicates the balancing weights. Represents a node and nodes The similarity between them.

[0034] In step S14, based on the obtained matching similarity between nodes, two nodes with matching similarity greater than a preset threshold are selected and merged to obtain a new node.

[0035] In step S15, after the merging is completed, the edges that were connected to the original nodes are connected to the new nodes.

[0036] In this invention, when performing entity recognition and merging, the entity to be identified and its corresponding attributes in the multidimensional polymorphic digital architecture data can be obtained. Based on the entity to be identified and its corresponding attributes, a node graph can be constructed, and multiple nodes in the node graph can be merged with each other. After obtaining the node graph, the similarity between two nodes can be calculated based on the attributes in the node graph, and these two nodes can be nodes to be matched. The neighbor set corresponding to the node is obtained, and then the semantic path similarity between the two nodes can be calculated using formula (1). The neighbor set can be other data nodes connected to the node. Based on the obtained similarity between nodes and the semantic path similarity between nodes, the matching similarity between nodes can be calculated using formula (2). According to the obtained matching similarity between nodes, two nodes with a matching similarity greater than a preset threshold can be filtered, and the two nodes can be merged to obtain a new node. After merging, the edge connected to the original node can be connected to the new node.

[0037] In one embodiment of the present invention, the process of calculating the similarity between two nodes may include: In step S16, the two nodes to be matched are obtained, and the attributes of the corresponding nodes are obtained.

[0038] In step S17, the similarity between two nodes is calculated using formula (3) based on the obtained attributes: Formula (3), in, Indicates the total number of attributes. Represents attributes The weight, This indicates an indicator function, which has attributes in both nodes. The value is 1 when it is not empty, and 0 in all other cases. Indicates a single attribute The similarity between the two nodes.

[0039] In this invention, when solving the similarity of nodes based on the attributes corresponding to the nodes, the similarity can be solved based on the matching degree of the attributes, and the importance of the attributes can also be reflected in the formula (3). The weight of the main attributes will be greater, so the similarity of nodes can be solved based on the important attributes.

[0040] In one embodiment of the present invention, such as Figure 4 As shown, the relationship fusion process may include: In step S18, the identified entities and their corresponding attributes, as well as various relationships between the entities and the entities corresponding to the attributes, are obtained.

[0041] In step S19, entities and various relationships between entities are fed into a relationship ranking model to rank the relationship levels between entities.

[0042] In step S20, after the sorting is completed, it is determined whether there are any conflicting relationships between the entities in the sorting.

[0043] In step S21, when there is a conflict between the relationships between entities in the sorting process, the relationships between entities that are ranked later are discarded.

[0044] In step S22, the relationships between the sorted entities are weighted, then merged, and the relationships between the entities that are ranked first are highlighted.

[0045] In this invention, during relationship fusion, identified entities and their corresponding attributes, as well as various relationships between entities and their corresponding attributes, can be obtained. After obtaining the relationships between entities, the entities and their various relationships can be fed into a relationship ranking model, thereby ranking the relationship levels between entities. After ranking, it can be determined whether there are conflicting relationships between the ranked entities. If there are conflicting relationships between the ranked entities, the relationships between entities ranked lower are discarded. Through the above processing, the ranked relationships can be streamlined, thereby avoiding the existence of conflicting relationships. After streamlining, the ranked relationships between entities can be weighted and then fused, highlighting the relationships between entities ranked higher.

[0046] In one embodiment of the present invention, such as Figure 5 As shown, the process of ranking can include: In step S23, the relational ranking model calculates the ranking score based on the relationships between entities using formula (4): Formula (4), in, Indicates the sorting score. Positional parameters representing the relationships between entities in the sorting process. This indicates the number of relationships between entities in the sorting. Indicates the first The correlation between relationships between entities.

[0047] In step S24, the relationships between entities are sorted in an ideal order, and their corresponding parameters are imported into formula (4) to obtain the ideal sorting score.

[0048] In step S25, the normalized cumulative gain of loss is calculated using formula (5) based on the obtained ideal ranking score: Formula (5), in, This represents the normalized cumulative gain. This represents the score for the ideal sorting.

[0049] In step S26, the goal is to maximize the cumulative gain of normalized loss in order to obtain the trained relational ranking model.

[0050] In step S27, the entities that need to be sorted and the various relationships between entities are fed into the trained relationship ranking model to sort the relationship levels between entities.

[0051] In this invention, during the sorting process, the relational ranking model can calculate the ranking score based on the relationships between entities using formula (4). Then, the relationships between entities can be sorted according to an ideal order, and their corresponding parameters can be imported into formula (4) to obtain the ideal ranking score. Based on the obtained ideal ranking score, the normalized depreciation cumulative gain can be calculated using formula (5). After obtaining the normalized depreciation cumulative gain, the model can be trained with the goal of maximizing the normalized depreciation cumulative gain, thereby obtaining the trained relational ranking model. After training, the entities to be sorted and the various relationships between entities can be fed into the trained relational ranking model, thereby allowing the ranking of the relationships between entities to be sorted.

[0052] In one embodiment of the present invention, such as Figure 6 As shown, the process for obtaining correlations may include: In step S28, the confidence level, reliability, timeliness, and number of pieces of evidence of the source of the relationship between entities are obtained.

[0053] In step S29, the relevance of the relationships between entities is classified into levels such as unrelated, somewhat related, related, highly related, very related, and perfectly related, thereby obtaining a level matrix.

[0054] In step S30, an initial association matrix is ​​obtained based on the confidence level, reliability, timeliness, and membership degree of the source of the obtained inter-entity relationship and the number of evidences and the level classification.

[0055] In step S31, the obtained initial correlation matrix is ​​set together to obtain a fuzzy relation matrix.

[0056] In step S32, weights are assigned to the confidence, reliability, timeliness, and number of evidence of the source of the relationship between entities to obtain a weight matrix.

[0057] In step S33, the weight matrix is ​​multiplied by the fuzzy relation matrix to obtain the comprehensive correlation matrix.

[0058] In step S34, the obtained comprehensive association matrix is ​​matched with the rank matrix to obtain the rank of the relationship between the corresponding entities, thereby obtaining the corresponding relevance.

[0059] In this invention, when acquiring relevance, data such as the confidence level, reliability, timeliness, and number of evidence sources of inter-entity relationships can be obtained. Then, the relevance of inter-entity relationships can be categorized into levels such as irrelevant, somewhat relevant, relevant, highly relevant, very relevant, and perfectly relevant, thus obtaining a level matrix. After the categorization, an initial association matrix can be obtained based on the membership degree between the acquired confidence level, reliability, timeliness, and number of evidence sources of inter-entity relationships and the levels of the categorization. After obtaining the association matrix, the initial association matrices can be aggregated to obtain a fuzzy relationship matrix. Weights are assigned to the confidence level, reliability, timeliness, and number of evidence sources of inter-entity relationships to obtain a weight matrix. This weight matrix is ​​then multiplied by the fuzzy relationship matrix to obtain a comprehensive association matrix. Based on the obtained comprehensive association matrix, it can be matched with the level matrix to obtain the relevance level of the corresponding inter-entity relationship, and further, the corresponding relevance can be obtained.

[0060] In one embodiment of the present invention, the multidimensional polymorphic digital architecture data includes structured, semi-structured, and unstructured data. Structured data is stored in business process systems, application servers, and database management system platforms, while semi-structured and unstructured data are scattered in various documents, such as project requirements specifications, feasibility study reports, and preliminary designs, which contain a wealth of data information.

[0061] On the other hand, the present invention can also provide a system for constructing a graph database based on a multidimensional polymorphic digital architecture. This system may include a data acquisition module and a construction module. The data acquisition module can be used to acquire multidimensional polymorphic digital architecture data. The construction module can be used to execute the method for constructing a graph database based on the acquired multidimensional polymorphic digital architecture, as described above.

[0062] Through the above technical solution, the method and system for constructing a graph database based on a multidimensional polymorphic digital architecture provided by this invention acquires multidimensional polymorphic digital architecture data and then preprocesses the acquired data. After preprocessing, entity identification and merging can be performed on the multidimensional polymorphic digital architecture data, thereby merging some duplicate entities. Simultaneously with entity identification and merging, attribute fusion can be performed on the identified entities. This attribute fusion can be based on the corresponding attributes of the identified entities or other methods. After entity identification and attribute fusion, the relationships between entities can be fused based on the entity identification and attribute fusion. After fusion, a graph database can be constructed based on the entities resulting from attribute fusion and relationship fusion. After the database is constructed, it can be optimized and updated periodically as needed. This method can collect data on a large number of architectural assets, perform relationship fusion, and then construct an intuitive node graph to visually display the relationships between various architectural assets.

[0063] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0064] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0065] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1The function specified in one or more boxes.

[0066] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0067] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0068] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, like read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0069] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0070] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0071] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for constructing a graph database based on a multidimensional polymorphic digital architecture, characterized in that, The method includes: Acquire multidimensional and polymorphic digital architecture data and preprocess the acquired data; After preprocessing, entity recognition and merging are performed on the multidimensional polymorphic digital architecture data; While performing entity identification and merging on the multidimensional and polymorphic digital architecture data, attribute fusion is performed on the identified entities; The relationships between the entities are fused based on entity recognition and attribute fusion; A graph database is constructed based on entities obtained through attribute fusion and relationship fusion, and the constructed database is optimized and updated regularly.

2. The method according to claim 1, characterized in that, Acquire multidimensional and polymorphic digital architecture data, and preprocess the acquired data, including: Acquire the multidimensional polymorphic digital architecture data and perform data cleaning on the multidimensional polymorphic digital architecture data; After cleaning, the relational tables, XML files, and JSON objects in the cleaned multidimensional polymorphic digital architecture data are converted into node and edge structures suitable for graph databases; The numerical data in the multidimensional polymorphic digital architecture data are normalized.

3. The method according to claim 1, characterized in that, After preprocessing, entity identification and merging are performed on the multidimensional and polymorphic digital architecture data, including: Obtain the entities to be identified and their corresponding attributes from the multidimensional polymorphic digital architecture data; Based on the entities to be identified and their corresponding attributes, construct a node graph; The similarity between two nodes is calculated based on the attributes in the node graph. Obtain the neighbor set corresponding to the node, and calculate the semantic path similarity between the two nodes using formula (1): Official (1), in, Represents a node and nodes The similarity of semantic paths between them Representing a path The weight, Indicates the connection node and nodes The set of semantic paths Represents a node The degree, Represents a node The degree; Based on the obtained similarity between nodes and the similarity of semantic paths between nodes, the matching similarity between nodes is calculated using formula (2): Official (2), in, Represents a node and nodes Matching similarity, Indicates the balancing weights. Represents a node and nodes The similarity between them; Based on the obtained matching similarity between nodes, two nodes with a matching similarity greater than a preset threshold are selected and merged to obtain a new node; After the merge is complete, the edges that were connected to the original nodes are connected to the new nodes.

4. The method according to claim 3, characterized in that, Calculating the similarity between two nodes based on attributes in the node graph includes: Get the two nodes to be matched, and get the attributes of the corresponding nodes; The similarity between two nodes is calculated using formula (3) based on the obtained attributes: Official (3), in, Indicates the total number of attributes. Representing attributes The weight, This indicates an indicator function, which has attributes in both nodes. The value is 1 when it is not empty, and 0 in all other cases. Indicates a single attribute The similarity between the two nodes.

5. The method according to claim 1, characterized in that, Based on entity recognition and attribute fusion, the relationships between the entities are fused, including: Obtain the identified entities and their corresponding attributes, as well as the various relationships between the entities and their corresponding attributes; The entities and the various relationships between them are fed into a relationship ranking model to rank the relationship levels between the entities. After sorting, determine whether there are any conflicting relationships between the entities in the sorting; When there are conflicting relationships between entities in the sorting process, discard the relationships between entities that are ranked lower. The relationships between the sorted entities are weighted, then merged, and the relationships between the entities that are ranked first are highlighted.

6. The method according to claim 5, characterized in that, The entities and various relationships between them are fed into a relationship ranking model to rank the relationships between entities, including: The relationship ranking model calculates the ranking score based on the relationships between entities using formula (4): Official (4), in, Indicates the sorting score. Positional parameters representing the relationships between entities in the sorting process. This indicates the number of relationships between entities in the sorting. Indicates the first The relevance of relationships between entities; The relationships between entities are sorted in an ideal order, and their corresponding parameters are imported into the formula (4) to obtain the ideal sorting score. The normalized cumulative gain of loss is calculated using formula (5) based on the obtained ideal ranking score: Official (5), in, This represents the normalized cumulative gain. Indicates the score of the ideal sort; The goal is to maximize the cumulative gain of the normalized loss in order to obtain the trained relational ranking model. The entities that need to be sorted and the various relationships between them are fed into the trained relationship ranking model to sort the relationship levels between the entities.

7. The method according to claim 6, characterized in that, To obtain the relevance of relationships between entities, including: The confidence level, reliability, timeliness, and number of pieces of evidence for obtaining the source of the relationship between entities; The relevance of relationships between entities is classified into levels such as unrelated, somewhat related, related, highly related, very related, and perfectly related, thus obtaining a level matrix; An initial association matrix is ​​obtained based on the confidence, reliability, timeliness, and membership degree of the source of the relationships between entities and the number of evidence and the level of the classification. The obtained initial correlation matrices are aggregated to obtain a fuzzy relation matrix; Weights are assigned to the confidence level, reliability, timeliness, and number of evidence of the source of the relationship between entities to obtain a weight matrix; The weight matrix is ​​multiplied by the fuzzy relation matrix to obtain the comprehensive correlation matrix; Based on the obtained comprehensive association matrix, it is matched with the level matrix to obtain the level of relevance between the corresponding entities, thereby obtaining the corresponding relevance.

8. The method according to claim 1, characterized in that, The multidimensional and polymorphic digital architecture data includes structured, semi-structured, and unstructured data. Structured data is stored in business process systems, application servers, and database management system platforms, while semi-structured and unstructured data are scattered in various documents, such as project requirements specifications, feasibility study reports, and preliminary designs, which contain a wealth of data information.

9. A system for constructing a graph database based on a multidimensional polymorphic digital architecture, characterized in that, The system includes: The data acquisition module is used to acquire multidimensional and polymorphic digital architecture data; A construction module is used to construct a graph database based on the acquired multidimensional polymorphic digital architecture data and execute the method for constructing a graph database based on the multidimensional polymorphic digital architecture as described in any one of claims 1-8.

Images