Enterprise digital architecture intelligent mapping implementation method and system

Abstract

The embodiment of the invention provides an enterprise digital architecture intelligent mapping implementation method and system, and belongs to the technical field of enterprise architecture mapping. The method comprises the following steps of: carding and converting contents of an architecture design result, an architecture design standard specification and an architecture management and control rule by adopting different conversion strategies according to an architecture asset type of a company to form structured and unstructured conversion results; performing multi-dimensional label management on the conversion result of the architecture assets, and realizing storage and warehousing of the architecture assets with label attributes in combination with a knowledge graph technology to form an architecture achievement standard library; and for structured, semi-structured and non-structured data in the architecture achievement standard library, performing slicing and index establishment based on an RAG technology to form an architecture knowledge base for digital architecture design and management and control. According to the method, enterprise architectures can be comprehensively collected, and intelligent mapping can be realized according to needs.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise architecture mapping, and in particular to a method and system for realizing intelligent mapping of enterprise digital architecture. Background Art

[0002] As the demand for digital construction and projects "doubles", the preparation of project feasibility studies, general design and other plans requires a lot of manpower and material resources, and the traditional way of relying on manual experience to carry out digital system architecture design and control is difficult to meet the requirements of high-standard architecture design, agile implementation of business needs, and rigid execution of architecture requirements, which affects the quality of digital transformation to a certain extent. In the existing digital architecture management, different departments and units have inconsistent expressions and understandings of architecture terms, views, specifications, solutions, etc., and lack a set of structured, standardized, intuitive and easy-to-use architecture design models. Therefore, a method for realizing intelligent mapping of enterprise digital architecture is needed to intuitively display the enterprise's architecture. Summary of the invention

[0003] The purpose of the embodiments of the present invention is to provide a method and system for realizing intelligent mapping of enterprise digital architecture, which can comprehensively collect enterprise architecture and realize intelligent mapping as needed.

[0004] In order to achieve the above-mentioned object, an embodiment of the present invention provides a method for realizing intelligent mapping of enterprise digital architecture, the method comprising:

[0005] According to the company's architecture asset type, different conversion strategies are used to sort out and convert the architecture design results, architecture design standards and specifications, and architecture management and control rules to form structured and unstructured conversion results;

[0006] Carry out multi-dimensional tag management on the transformation results of architecture assets, combine knowledge graph technology to realize the storage of architecture assets with tag attributes, and form a standard library of architecture achievements;

[0007] For the structured, semi-structured and unstructured data in the standard library of architecture achievements, slice and index them based on RAG technology to form an architecture knowledge base for digital architecture design and management;

[0008] Building a knowledge retrieval engine based on RAG and the architecture knowledge base;

[0009] The knowledge retrieval engine and GPT technology are integrated with the enterprise architecture rule base, unstructured architecture assets and structured architecture assets to achieve intelligent mapping of the enterprise digital architecture based on user needs.

[0010] Optionally, the architecture design achievements include: business architecture achievements, application architecture achievements, data architecture achievements, technical architecture achievements, and security architecture achievements;

[0011] For the business architecture results, we sorted out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Business Architecture Volume" to extract the relevant business functions, processes, organizational units, relationships and cross-professional business collaboration relationships of the enterprise middle office;

[0012] For the application architecture results, we sorted out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Application Architecture Volume" to extract the system functions, boundaries, integration relationships based on the construction of the enterprise middle platform, and the correspondence between functions and business functions;

[0013] For the data architecture results, we sorted out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Data Architecture Volume", extracted the company's unified data model, and the flow and distribution of data entities in the data middle platform among various current application functions;

[0014] For the technical architecture results, the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Technical Architecture Volume" was sorted out to extract the underlying supporting technical components and service capabilities such as the technical middle platform, State Grid Cloud, IoT management platform, data middle platform, business middle platform and enterprise-level real-time measurement center;

[0015] For the security architecture results, we sorted out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Security Architecture Volume" and extracted the technical solutions and rules related to the security protection of the State Grid Cloud and the enterprise middle platform;

[0016] The corresponding design results are structurally transformed in the form of "asset name" + "asset attributes" + "asset relationship".

[0017] Optionally, sorting out and transforming the architecture design standard specification includes:

[0018] Sort out the "Framework Design of Digital Technology Support System for New Power System", "Technical Management Measures for Grid Digitalization Projects of State Grid Corporation of China", the digital architecture planning and design of State Grid Corporation of China, the digital architecture standard design of State Grid Corporation of China, the SG-ERP architecture design of State Grid Corporation of China, the top-level design of the State Grid cloud platform, the technical solution for cloud migration of business systems, the top-level design of the enterprise middle platform, and the artificial intelligence platform architecture design solution, and design structured transformation according to the content definition; for texts that lack causal hierarchical relationships, directly perform unstructured transformation according to the file format.

[0019] Optionally, sorting out and transforming the architecture control rules includes:

[0020] We sorted out the architecture management and control process and rule documents such as the "Management Measures for Digital Construction of Power Grids of State Grid Corporation of China", the "Management Measures for Information Architecture of State Grid Corporation of China", and the Implementation Rules for Digital Architecture Management of State Grid Corporation of China, extracted the review points, and performed item-by-item decomposition and quantification based on text intelligence technology, and mapped them according to the language logic of the rules; for some review points that could not be logically mapped, they were directly converted into unstructured form.

[0021] Optionally, multi-dimensional tag management is performed on the transformation results of architecture assets. Combined with knowledge graph technology, architecture assets with tag attributes can be stored in the library to form a standard library of architecture achievements including:

[0022] Obtain the architecture design results, design specifications, and control points obtained after the structured transformation of digital architecture assets;

[0023] For the architecture design results, design specifications, and key points of control, label management can be performed according to the architecture category, project construction form, project stage, architecture status, control importance level, the type of middle platform involved in the design rules, and component type; the architecture categories include: business, application, data, technology, and security architecture; the project construction forms include: integrated construction and self-built projects; the project stages include: feasibility study, general design, implementation, launch, and acceptance; the architecture status includes the current status and the design status;

[0024] The results of the structured transformation and label management of digital architecture assets are stored in a graph database using knowledge graph technology, and the hierarchical structure and dependencies of the architecture elements are displayed to form a standard library of results.

[0025] Optionally, for the structured, semi-structured and unstructured data in the standard library of architecture achievements, slices and indexes are established based on RAG technology to form an architecture knowledge base for digital architecture design and management, including:

[0026] Obtain structured, semi-structured and unstructured data in the achievement standard library;

[0027] Extract key information from structured, semi-structured and unstructured data through RAG;

[0028] The key information in different types of structured, semi-structured and unstructured data is cleaned, format converted and standardized to merge, and the asset entity data with the same attributes are aligned;

[0029] Through natural language processing tasks, the various types of fused data are converted into low-dimensional dense vectors, so that the data can capture the meaningful structure and relationship of the original data in the new vector space, classify the identified entities, and establish a relationship network between them to form a structured knowledge representation;

[0030] Use RAG to complete the knowledge of the converted data and remove noise, HTML tags, special characters and stop words in the text;

[0031] The processed text contents are concatenated to form a complete text string, and the text string is organized and cut to obtain a list of text paragraphs to form an architecture knowledge base.

[0032] Optionally, the knowledge retrieval engine and the GPT technology are integrated with the enterprise architecture rule base, unstructured architecture assets, and structured architecture assets to realize intelligent mapping of the enterprise digital architecture according to user needs, including:

[0033] Fusion of the knowledge retrieval engine with heterogeneous data of GPT;

[0034] The fused knowledge retrieval engine retrieves the most relevant information fragments to the input query from multiple heterogeneous data sources, and quickly locates the part of the data fragment that is most likely to contain the answer through keyword matching and vector similarity calculation methods;

[0035] The knowledge retrieval engine organizes relevant information fragments after preliminary processing of the retrieved information to form a comprehensive context input;

[0036] The fused GPT understands the retrieved information based on the contextual information provided by the knowledge retrieval engine;

[0037] The GPT fuses the retrieved discrete pieces of information into a logically consistent, information-rich answer or narrative.

[0038] Optionally, the knowledge retrieval engine and GPT technology are trained to integrate the enterprise architecture rule base, unstructured architecture assets, and structured architecture assets to achieve intelligent mapping of enterprise digital architecture according to user needs, including:

[0039] The user selects a template for the diagram, an architectural asset, which defines the basic structure of the view, element layout, and necessary labels and connection relationships;

[0040] The fused GPT generates an architecture view based on the retrieved answer or narrative, the user-selected template and the architecture assets, and the architecture view includes asset names, relationships between assets, and asset descriptions.

[0041] In another aspect, the present invention provides a system for realizing intelligent mapping of enterprise digital architecture, the system comprising:

[0042] User input module, used to select system or project information according to requirements;

[0043] The intelligent mapping module is used to execute the above-mentioned method for realizing intelligent mapping of enterprise digital architecture according to user needs.

[0044] Through the above technical scheme, the present invention provides a method and system for realizing intelligent mapping of enterprise digital architecture. According to the company's architecture asset type, different conversion strategies are used to sort out and convert the contents of architecture design results, architecture design standards and specifications, and architecture management and control rules, so that structured and unstructured conversion results can be formed. The transformation results of architecture assets can be managed by multi-dimensional tags, and combined with knowledge graph technology, architecture assets with label attributes can be stored in the library, so that an architecture achievement standard library can be formed. For structured, semi-structured and unstructured data in the architecture achievement standard library, slicing and indexing can be established based on RAG technology, so that an architecture knowledge base for digital architecture design and management can be formed. A knowledge retrieval engine can be constructed based on RAG and the architecture knowledge base. The knowledge retrieval engine and GPT technology are integrated with the enterprise architecture rule base, unstructured architecture assets and structured architecture assets, so that the intelligent mapping of enterprise digital architecture can be realized according to user needs. This method can comprehensively collect enterprise architecture and realize intelligent mapping as needed.

[0045] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

[0047] Figure 1 It is a flow chart of a method for realizing intelligent mapping of enterprise digital architecture according to an embodiment of the present invention;

[0048] Figure 2 It is a flowchart of converting design results of a method for realizing intelligent mapping of enterprise digital architecture in one embodiment of the present invention;

[0049] Figure 3 It is a flow chart of forming an architecture achievement standard library of an enterprise digital architecture intelligent mapping implementation method according to an embodiment of the present invention;

[0050] Figure 4It is a flowchart of forming an architecture knowledge base of a method for realizing intelligent mapping of enterprise digital architecture according to one embodiment of the present invention;

[0051] Figure 5 It is a first flow chart of a method for realizing intelligent mapping of an enterprise digital architecture according to an embodiment of the present invention;

[0052] Figure 6 It is a second flow chart of a method for realizing intelligent mapping of an enterprise digital architecture according to an embodiment of the present invention;

[0053] Figure 7 It is an example code of template input information for implementing intelligent mapping of a method for implementing intelligent mapping of enterprise digital architecture according to an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.

[0055] Figure 1 The flowchart of a method for realizing intelligent mapping of enterprise digital architecture according to an embodiment of the present invention is shown in FIG. In the present invention, the process of the method for realizing intelligent mapping may include:

[0056] In step S1, according to the company's architectural asset type, different conversion strategies are used to sort out and convert the contents of architectural design results, architectural design standards and specifications, and architectural management and control rules to form structured and unstructured conversion results.

[0057] In step S2, the transformation results of the architecture assets are managed with multi-dimensional labels, and the architecture assets with label attributes are stored in the library in combination with knowledge graph technology to form a standard library of architecture achievements.

[0058] In step S3, the structured, semi-structured and unstructured data in the standard library of architectural achievements are sliced ​​and indexed based on RAG technology to form an architectural knowledge base for digital architecture design and management.

[0059] In step S4, a knowledge retrieval engine is constructed based on the RAG and the architecture knowledge base.

[0060] In step S5, the knowledge retrieval engine and GPT technology are integrated with the enterprise architecture rule base, unstructured architecture assets and structured architecture assets to realize intelligent mapping of the enterprise digital architecture according to user needs.

[0061] In the present invention, when performing intelligent mapping, different conversion strategies can be used to sort out and convert the contents of architecture design results, architecture design standards and specifications, and architecture management and control rules according to the company's architecture type, so that structured and unstructured conversion results can be formed, and then the transformation results of the architecture assets can be managed by multi-dimensional tags, combined with knowledge graph technology, so that the architecture assets with label attributes can be stored in the library to form an architecture achievement standard library. For the structured, semi-structured and unstructured data in the architecture achievement standard library, slicing and indexing can be performed based on RAG technology, so that an architecture knowledge base for digital architecture design and management can be formed, and then a knowledge retrieval engine can be constructed based on RAG and the architecture knowledge base. The knowledge retrieval engine and GPT technology are integrated with the enterprise architecture rule base, unstructured architecture assets and structured assets, so that the intelligent mapping of the enterprise digital architecture can be realized according to user needs. This method can comprehensively collect enterprise architecture and realize intelligent mapping as needed.

[0062] In one embodiment of the present invention, Figure 2 As shown in the figure, the process of transforming design results may include:

[0063] In step S6, for the business architecture results, sort out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Business Architecture Volume" to extract the relevant business functions, processes, organizational units, relationships and cross-professional business collaboration relationships of the enterprise middle platform.

[0064] In step S7, for the application architecture results, sort out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Application Architecture Volume" to extract the system functions, boundaries, integration relationships based on the enterprise middle platform construction, and the correspondence between functions and business functions.

[0065] In step S8, for the data architecture results, sort out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Data Architecture Volume", extract the company's unified data model, and the flow distribution of data entities in the data middle platform among various current application functions.

[0066] In step S9, for the technical architecture results, sort out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Technical Architecture Volume" to extract the underlying supporting technical components and service capabilities such as the technical middle platform, State Grid Cloud, Internet of Things management platform, data middle platform, business middle platform and enterprise-level real-time measurement center.

[0067] In step S10, for the security architecture results, sort out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Security Architecture Volume" to extract the technical solution rules related to the security protection of the State Grid Cloud and the enterprise middle platform.

[0068] In step S11, the corresponding design results are structurally transformed in the form of "asset name" + "asset attribute" + "asset relationship".

[0069] Architecture assets may include architecture status, architecture blueprint, architecture evolution path, architecture supporting assets, architecture principles, management and control methods, and architecture specifications. According to the type characteristics of architecture assets, different methods are used to structure and transform them to form a unified standard data form. In the present invention, architecture design results may include: business architecture results, application architecture results, data architecture results, technical architecture results, and security architecture results. Based on the architecture design results, systematic extraction and disassembly can be carried out, mainly in a "structured" form, so as to form quantifiable structured assets.

[0070] Business architecture results: Sort out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Business Architecture Volume", extract the relevant business functions, processes, organizational units, relationships and cross-professional business collaboration relationships of the enterprise middle platform.

[0071] Application architecture results: sort out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Application Architecture Volume" to extract the system functions, boundaries, integration relationships based on the enterprise middle platform construction, as well as the correspondence between functions and business functions.

[0072] Data architecture results: Sort out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Data Architecture Volume", extract the company's unified data model, and the flow and distribution of data entities in the data middle platform among various current application functions.

[0073] Technical architecture results: sort out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Technical Architecture Volume" to extract the underlying supporting technical components and service capabilities such as technical middle platform, State Grid Cloud, Internet of Things management platform, data middle platform, business middle platform and enterprise-level real-time measurement center.

[0074] Security architecture results: Sort out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Security Architecture Volume" and extract the technical solution rules related to the security protection of the State Grid Cloud and the enterprise middle office.

[0075] After sorting out the results of the architecture design, the corresponding design results can be structurally transformed in the form of "asset name" + "asset attribute" + "asset relationship". For example, a business process can be defined in dimensions such as process name (asset name), upper and lower process nodes (asset relationship), and node business logic description (asset attribute).

[0076] In one embodiment of the present invention, when combing and transforming the architecture design standards and specifications, the "Framework Design of the New Power System Digital Technology Support System", "Technical Management Measures for the Digitalization Project of State Grid Corporation of China", the digital architecture planning and design of State Grid Corporation, the digital architecture standard design of State Grid Corporation, the SG-ERP architecture design of State Grid Corporation, the top-level design of the State Grid cloud platform, the technical solution for clouding business systems, the top-level design of the enterprise middle platform, the architecture design of the artificial intelligence platform and other top-level design solutions can be combed, and the structured transformation method can be designed according to the content definition, such as the system design view, and the "architecture elements" and "element relationships" in the view can be extracted to perform structured design standards and requirements; for design specifications and design standards in text form, NLP technology can be used to perform item-based disassembly of the specification content to achieve content quantification, and then the processed files can be structured based on the text form. For texts that lack causal hierarchical relationships, or other undisassembled content such as extracted pictures, unstructured transformation can be performed directly according to the file format.

[0077] In the present invention, when combing and transforming the architecture control rules, the relevant content is extracted and disassembled according to the content characteristics of the architecture control rules, such as architecture review requirements, architecture control templates, etc., and the transformation is mainly carried out in the form of "structured + unstructured". The architecture control process and rule documents such as the "Management Measures for Digital Construction of Power Grids of State Grid Corporation of China", "Management Measures for Information Architecture of State Grid Corporation of China", and the Implementation Rules for Digital Architecture Management of State Grid Corporation of China are sorted out, and the review points are extracted, and the item-based disassembly and quantification are carried out based on text intelligence technology, and mapped according to the language logic of the rules. Some review points cannot be logically mapped and are directly transformed in an unstructured form.

[0078] In one embodiment of the present invention, Figure 3 As shown, the process of forming a standard library of architectural achievements may include:

[0079] In step S12, the architecture design results, design specifications, and control points obtained after the structured transformation of the digital architecture assets are obtained.

[0080] In step S13, the architecture design results, design specifications, and management and control points can be labeled and managed according to the architecture category, project construction form, project stage, architecture status, management and control importance level, the type of middle platform involved in the design rules, and the component type; the architecture categories include: business, application, data, technology, and security architecture; the project construction forms include: integrated construction and self-built projects; the project stages include: feasibility study, conceptual design, implementation, launch, and acceptance; the architecture status includes the current status and the design status.

[0081] In step S14, the results of the structured transformation of digital architecture assets and the results of label management are stored in a graph database using knowledge graph technology, and the hierarchical structure and dependency relationships of the architecture elements are displayed to form a standard library of results.

[0082] In the present invention, when constructing the standard library of architectural achievements, the architectural design achievements, design specifications and key points of control obtained after the structured transformation of digital architectural assets can be obtained, and the labels can be maintained and classified according to multiple dimensions. By using the knowledge graph, the relationship between different labels and entities can be modeled, for example, the attributes, associations, hierarchies and other information between labels and architectural assets can be organically combined, so that entities with multi-dimensional label information can be more accurately found and managed, the efficiency of information retrieval and management can be improved, and more intelligent recommendations and data analysis can be supported. For architectural design achievements, design specifications and key points of control, label management can be performed according to architectural identification, project construction form, project stage, architectural status, control importance level, middle platform type involved in design rules, and component type. Architecture categories include: business, application, data, technology, and security architecture. Project construction forms include: integrated construction and self-built projects. Project stages include: feasibility study, general design, implementation, online, and acceptance. Architecture status includes current status and design status. For the results of the structured transformation of digital architecture assets and the results of label management, the graph database can be used as a carrier to store them using knowledge graph technology, and the hierarchical architecture and dependencies of its architecture elements can be displayed to form a standard library of results. For the architecture design and control rule data, the results of its quantification and structured processing are combined with multi-dimensional label annotations to form an architecture rule library, thereby achieving comprehensive, efficient, accurate, and systematic control and guidance of the architecture result library data, and continuously improving data quality. Multi-dimensional labels can be used to achieve accurate retrieval and positioning, and support the rapid acquisition of architecture assets. The architecture result standard library is associated with the architecture rule library to achieve a perfect connection between architecture design and control work. For example, in the process of architecture review, the review points are retrieved according to the labels, and the architecture design plan of the current project is compared to automatically determine whether it meets the rule requirements.

[0083] In one embodiment of the present invention, Figure 4 As shown, the process of forming an architecture knowledge base may include:

[0084] In step S15, structured, semi-structured and unstructured data in the achievement standard library are obtained.

[0085] In step S16, key information in structured, semi-structured and unstructured data is extracted through RAG.

[0086] In step S17, key information in different types of structured, semi-structured and unstructured data is cleaned, format converted and standardized for integration, and asset entity data with the same attributes are aligned.

[0087] In step S18, the various types of fused data are converted into low-dimensional dense vectors through natural language processing tasks, so that the data can capture the meaningful structure and relationship of the original data in the new vector space, classify the identified entities, and establish a relationship network between them to form a structured knowledge representation.

[0088] In step S19, RAG is used to complete the converted data and remove noise, HTML tags, special characters and stop words in the text.

[0089] In step S20, the processed text contents are concatenated to form a complete text string, and the text string is organized and cut to obtain a text paragraph list to form an architecture knowledge base.

[0090] In the present invention, when constructing an architecture knowledge base, structured, semi-structured and unstructured data in the achievement standard library can be obtained. Among them, unstructured data can include: scientific research, business needs, software needs, outline design, security protection solutions, etc. Semi-structured data can include: Excel data import, Json data parsing, etc. Structured data can include: business architecture assets, application architecture assets, data architecture assets, technical architecture assets, security architecture assets, etc. Converted into an operable text format through intelligent parsing technology. At the same time, structured architecture assets are collected to ensure that all data are ready. After obtaining the data, key information in structured, semi-structured and unstructured data can be extracted through RAG. This stage not only involves the classification and identification of assets, but also includes the identification and extraction of complex asset relationships to ensure that each piece of data can be correctly understood and classified. After extracting key information, the key information in different types of structured, semi-structured and unstructured data can be cleaned, format converted and standardized, and the asset entity data with the same attributes can be aligned. Based on the fused data, the technology of converting various types of fused data into low-dimensional dense vectors through natural language processing tasks. This allows the data to capture the meaningful structure and relationships of the original data in the new vector space, and through technologies such as RDF (Resource Description Framework), the identified entities are classified and a relationship network between them is established to form a structured knowledge representation. RGA can be used to complete the knowledge of the converted data and remove noise, HTML tags, special characters, and stop words in the text. The relevant information extraction technology in RAG can concatenate the processed text content to form a complete text string. Subsequently, these texts can be carefully organized and cut, and finally a neat and structured list of text paragraphs is generated, and an architecture knowledge base is constructed based on the results.

[0091] In one embodiment of the present invention, Figure 5 As shown, the first process of realizing intelligent mapping may include:

[0092] In step S21, the knowledge retrieval engine is integrated with the heterogeneous data of GPT.

[0093] In step S22, the fused knowledge retrieval engine retrieves the information fragments most relevant to the input query from multiple heterogeneous data sources, and quickly locates the part most likely to contain the answer from the data fragments through keyword matching and vector similarity calculation methods.

[0094] In step S23, the knowledge retrieval engine preliminarily processes the retrieved information and organizes the relevant information fragments to form a comprehensive context input.

[0095] In step S24, the fused GPT understands the retrieved information based on the context information provided by the knowledge retrieval engine.

[0096] In step S25, GPT fuses the retrieved discrete information fragments into a logically consistent, information-rich answer or narrative.

[0097] In the present invention, when performing intelligent mapping, the heterogeneous data of the knowledge retrieval engine and GPT can be fused, and then the fused knowledge retrieval engine can retrieve the information fragments most relevant to the input query from multiple heterogeneous data sources. The heterogeneous data source may include an architecture knowledge base. The most likely part of the data fragment can be quickly located through keyword matching and vector similarity calculation methods. The knowledge retrieval engine can organize the relevant information fragments after preliminary processing of the retrieved information to form a comprehensive context input. The fused GPT can understand the retrieved information based on the context information provided by the knowledge retrieval engine, and then the GPT can judge and fuse the retrieved discrete information into a logically consistent and information-rich answer or narrative.

[0098] In one embodiment of the present invention, Figure 6 As shown, the second process of realizing intelligent mapping may include:

[0099] In step S26, the user selects a template for drawing and an asset for architecture, where the template defines the basic structure of the view, element layout, and necessary labels and connection relationships.

[0100] In step S27, the fused GPT generates an architecture view based on the retrieved answer or narrative, the template selected by the user and the architecture assets. The architecture view includes the asset name, the relationship between assets, and the asset description.

[0101] In the present invention, when performing intelligent mapping, a template can be designed for each type of architecture view (such as a technical architecture view, a business architecture view, a network security topology map, etc.). Figure 7Enter information sample code for the automatic drawing template. The template defines the basic structure of the view, element layout, necessary tags and connection relationships, etc. The template can be in XML, JSON or other structured formats to facilitate program parsing and operation. Users can select the template for drawing and the assets of the architecture. Combined with the knowledge data of the RAG knowledge base such as a large number of typical case templates, architecture assets, typical design results, etc., extract key information and establish an index, and use TF-IDF, BM25 or more advanced semantic vector models to create a searchable knowledge base. GPT can generate an architecture view based on the fused information based on the retrieved answers or narratives, refer to the template for drawing and the assets of the architecture selected by the user, and use the GPT model. The view can include a graphical architecture view file (such as PNG, SVG), or structured view description information, including text information such as asset name, relationship between assets, asset description, etc. GPT can also perform a certain degree of creative generation and optimization based on the specific requirements of the State Grid or previous architecture design experience.

[0102] On the other hand, the present invention also provides a system for realizing intelligent mapping of enterprise digital architecture, the system comprising:

[0103] User input module, used to select system or project information according to requirements.

[0104] The intelligent mapping module is used to execute the above-mentioned method for realizing intelligent mapping of enterprise digital architecture according to user needs.

[0105] Through the above technical scheme, the present invention provides a method and system for realizing intelligent mapping of enterprise digital architecture. According to the company's architecture asset type, different conversion strategies are used to sort out and convert the contents of architecture design results, architecture design standards and specifications, and architecture management and control rules, so that structured and unstructured conversion results can be formed. The transformation results of architecture assets can be managed by multi-dimensional tags, and combined with knowledge graph technology, architecture assets with label attributes can be stored in the library, so that an architecture achievement standard library can be formed. For structured, semi-structured and unstructured data in the architecture achievement standard library, slicing and indexing can be established based on RAG technology, so that an architecture knowledge base for digital architecture design and management can be formed. A knowledge retrieval engine can be constructed based on RAG and the architecture knowledge base. The knowledge retrieval engine and GPT technology are integrated with the enterprise architecture rule base, unstructured architecture assets and structured architecture assets, so that the intelligent mapping of enterprise digital architecture can be realized according to user needs. This method can comprehensively collect enterprise architecture and realize intelligent mapping as needed.

[0106] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0107] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0108] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0109] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0110] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0111] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0112] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules 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 technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0113] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0114] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for realizing intelligent mapping of enterprise digital architecture, characterized in that: The method comprises: According to the company's architecture asset type, different conversion strategies are used to sort out and convert the architecture design results, architecture design standards and specifications, and architecture management and control rules to form structured and unstructured conversion results; Carry out multi-dimensional tag management on the transformation results of architecture assets, combine knowledge graph technology to realize the storage of architecture assets with tag attributes, and form a standard library of architecture achievements; For the structured, semi-structured and unstructured data in the standard library of architecture achievements, slice and index them based on RAG technology to form an architecture knowledge base for digital architecture design and management; Building a knowledge retrieval engine based on RAG and the architecture knowledge base; The knowledge retrieval engine and GPT technology are integrated with the enterprise architecture rule base, unstructured architecture assets and structured architecture assets to achieve intelligent mapping of the enterprise digital architecture based on user needs.

2. The method according to claim 1, characterized in that The architecture design achievements include: business architecture achievements, application architecture achievements, data architecture achievements, technical architecture achievements, and security architecture achievements; For the business architecture results, we sorted out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Business Architecture Volume" to extract the relevant business functions, processes, organizational units, relationships and cross-professional business collaboration relationships of the enterprise middle office; For the application architecture results, we sorted out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Application Architecture Volume" to extract the system functions, boundaries, integration relationships based on the construction of the enterprise middle platform, and the correspondence between functions and business functions; For the data architecture results, we sorted out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Data Architecture Volume", extracted the company's unified data model, and the flow and distribution of data entities in the data middle platform among various current application functions; For the technical architecture results, the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Technical Architecture Volume" was sorted out to extract the underlying supporting technical components and service capabilities such as the technical middle platform, State Grid Cloud, IoT management platform, data middle platform, business middle platform and enterprise-level real-time measurement center; For the security architecture results, we sorted out the "State Grid Corporation of China_Informatization SG-ERP Architecture Design-Security Architecture Volume" and extracted the technical solutions and rules related to the security protection of the State Grid Cloud and the enterprise middle platform; The corresponding design results are structurally transformed in the form of "asset name" + "asset attribute" + "asset relationship".

3. The method according to claim 1, wherein the characteristic value is that sorting out and transforming the architecture design standard specification comprises: Sort out the "Framework Design of Digital Technology Support System for New Power System", "Technical Management Measures for Grid Digitalization Projects of State Grid Corporation of China", the digital architecture planning and design of State Grid Corporation of China, the digital architecture standard design of State Grid Corporation of China, the SG-ERP architecture design of State Grid Corporation of China, the top-level design of the State Grid cloud platform, the technical solution for cloud migration of business systems, the top-level design of the enterprise middle platform, and the artificial intelligence platform architecture design solution, and design structured transformation according to the content definition; for texts that lack causal hierarchical relationships, directly perform unstructured transformation according to the file format.

4. The method according to claim 1, characterized in that: Sorting out and transforming the architecture control rules include: We sorted out the architecture management and control process and rule documents such as the "Management Measures for Digital Construction of Power Grids of State Grid Corporation of China", the "Management Measures for Information Architecture of State Grid Corporation of China", and the Implementation Rules for Digital Architecture Management of State Grid Corporation of China, extracted the review points, and performed item-by-item decomposition and quantification based on text intelligence technology, and mapped them according to the language logic of the rules; for some review points that could not be logically mapped, they were directly converted into unstructured form.

5. The method according to claim 1, characterized in that Multi-dimensional tag management is performed on the transformation results of architecture assets. Combined with knowledge graph technology, architecture assets with tag attributes are stored in the library to form a standard library of architecture results including: Obtain the architecture design results, design specifications, and control points obtained after the structured transformation of digital architecture assets; For the architecture design results, design specifications, and key points of control, label management can be performed according to the architecture category, project construction form, project stage, architecture status, control importance level, the type of middle platform involved in the design rules, and component type; the architecture categories include: business, application, data, technology, and security architecture; the project construction forms include: integrated construction and self-built projects; the project stages include: feasibility study, general design, implementation, launch, and acceptance; the architecture status includes the current status and the design status; The results of the structured transformation and label management of digital architecture assets are stored in a graph database using knowledge graph technology, and the hierarchical structure and dependencies of the architecture elements are displayed to form a standard library of results.

6. The method according to claim 5, characterized in that For the structured, semi-structured and unstructured data in the standard library of architecture achievements, slices and indexes are established based on RAG technology to form an architecture knowledge base for digital architecture design and management, including: Obtain structured, semi-structured and unstructured data in the achievement standard library; Extract key information from structured, semi-structured and unstructured data through RAG; The key information in different types of structured, semi-structured and unstructured data is cleaned, format converted and standardized to merge, and the asset entity data with the same attributes are aligned; Through natural language processing tasks, the various types of fused data are converted into low-dimensional dense vectors, so that the data can capture the meaningful structure and relationship of the original data in the new vector space, classify the identified entities, and establish a relationship network between them to form a structured knowledge representation; Use RAG to complete the knowledge of the converted data and remove noise, HTML tags, special characters and stop words in the text; The processed text contents are concatenated to form a complete text string, and the text string is organized and cut to obtain a list of text paragraphs to form an architecture knowledge base.

7. The method according to claim 6, characterized in that The knowledge retrieval engine and GPT technology are integrated into the enterprise architecture rule base, unstructured architecture assets and structured architecture assets to realize intelligent mapping of enterprise digital architecture according to user needs, including: Fusion of the knowledge retrieval engine with heterogeneous data of GPT; The fused knowledge retrieval engine retrieves the most relevant information fragments to the input query from multiple heterogeneous data sources, and quickly locates the part of the data fragment that is most likely to contain the answer through keyword matching and vector similarity calculation methods; The knowledge retrieval engine organizes relevant information fragments after preliminary processing of the retrieved information to form a comprehensive context input; The fused GPT understands the retrieved information based on the contextual information provided by the knowledge retrieval engine; The GPT fuses the retrieved discrete pieces of information into a logically consistent, information-rich answer or narrative.

8. The method according to claim 7, characterized in that The knowledge retrieval engine and GPT technology are trained to integrate the enterprise architecture rule base, unstructured architecture assets and structured architecture assets to realize intelligent mapping of enterprise digital architecture according to user needs, including: The user selects a template for the diagram, an architectural asset, which defines the basic structure of the view, element layout, and necessary labels and connection relationships; The fused GPT generates an architecture view based on the retrieved answer or narrative, the user-selected template and the architecture assets, and the architecture view includes asset names, relationships between assets, and asset descriptions.

9. An enterprise digital architecture intelligent mapping implementation system, characterized in that: The system comprises: User input module, used to select system or project information according to requirements; The intelligent mapping module is used to execute the enterprise digital architecture intelligent mapping implementation method as described in any one of claims 1-8 according to user needs.

Images