System and method for constructing business and data fusion knowledge based on enterprise architecture

Abstract

The invention discloses a system and a method for constructing business and data fusion knowledge based on an enterprise architecture, relates to the field of enterprise informatization, and solves the problem that related data cannot be well acquired in a business problem solving process due to the fact that existing business knowledge lacks fusion with data and cannot well correspond to the data. According to the method, a framework of business and data fusion knowledge is defined, a hierarchical structure, content composition and an expression mode of the business and data fusion knowledge are formed, a standard method for establishing business and data fusion from an enterprise architecture (4A) is constructed, various information is extracted from a business architecture (BA) and a data architecture (IA), and knowledge is automatically generated and enhanced according to the knowledge framework. When the data is used, the relationship between the data and the business can be clearly known, the data can analyze and solve the business problem, the system can quickly position the required data and provide the analysis method and result, and the efficiency of solving the business problem by using the data is improved.

Description

Technical Field

[0001] The present application relates to the field of enterprise informatization, and in particular to a system and method for building business and data fusion knowledge based on enterprise architecture. Background Art

[0002] In an era of rapid development of big models and AI Agent technology (i.e. artificial intelligence agent technology, which is an intelligent technology that integrates perception, analysis, decision-making and execution capabilities), high-quality data and business knowledge are the prerequisites for big models and AI Agent to play a role. In addition, those involving enterprise information processing are generally designed based on the enterprise architecture (4A).

[0003] At present, there is data managed by tools such as data middle platform on the market, and there is also business knowledge established through business manuals, business processes, and system documents. However, these business knowledge often cannot correspond well to the data, resulting in the inability to obtain relevant data in the process of solving business problems. The core business knowledge lacks integration with data, and there are several major problems:

[0004] First, most companies have conducted data governance and formed data assets. The understanding of data assets is generally through data dictionaries, which include business definitions, data descriptions, data standards and other information, but there is still a lack of information on how to use data to solve business problems.

[0005] Second, the mainstream sources of business knowledge on the market are business system specifications, business manuals, business processes, or documents manually written by business personnel. Business knowledge includes business definitions, business rules, etc., but lacks methods and ways to use data to analyze and judge business status.

[0006] Third, business and data are constantly changing, and knowledge from business documents, business processes, etc. is often not updated in a timely manner. There is a lack of methods to closely integrate business and data into knowledge and ways to update knowledge. Summary of the invention

[0007] In order to improve the above-mentioned deficiencies, the present application provides a system and method for building business and data fusion knowledge based on enterprise architecture.

[0008] This application provides a system and method for building business and data fusion knowledge based on enterprise architecture, using the following technical solutions:

[0009] A system and method for constructing business and data fusion knowledge based on enterprise architecture, including enterprise architecture (4A), business and data fusion knowledge framework module, information extraction module, knowledge generation module, and knowledge service module;

[0010] The enterprise architecture (4A) includes business architecture (BA) and data architecture (IA);

[0011] The business and data fusion knowledge framework module includes knowledge classification, knowledge classification, knowledge stratification, and knowledge content templates, wherein knowledge classification and knowledge stratification are respectively combined with the business concerns of the business architecture (BA), using the source information of business knowledge and data, corresponding to the business attribution classification of knowledge, and recording the level of knowledge; the knowledge content template is used to define the type and structure of knowledge content, and provide a standard template for the subsequent automatic generation of business and data fusion knowledge based on the enterprise architecture (4A);

[0012] The information extraction module extracts information required for generating knowledge from the business architecture (BA) and the data architecture (IA), and increases the relationship between business and data;

[0013] The knowledge generation module: extracts information from the business architecture (BA) and the data architecture (IA), and automatically generates declarative knowledge and procedural knowledge, while enhancing the knowledge;

[0014] The knowledge service module stores and updates declarative knowledge and procedural knowledge, and provides knowledge services to the outside world.

[0015] Preferably, the knowledge content template includes declarative knowledge and procedural knowledge, the declarative knowledge includes recording comprehensive business definitions, instructions, usage methods, and rules, and the procedural knowledge includes recording processes, steps, methods, and actions.

[0016] Preferably, the business architecture (BA) includes L1 (process classification), L2 (process group), L3 (business process), L4 (industry activities), and L5 (tasks), and the information extraction module collects business segments, business processes, business activities, and business entities therefrom, and summarizes them into business classification, business definition, business process, and business indicators.

[0017] Preferably, the data architecture (IA) includes L1 (data domain), L2 (subject group), L3 (data entity), L4 (logical entity), and L5 (attribute). The information extraction module collects information such as data entity, logical entity, attribute, etc. from the data architecture (IA), including data definition, data type, data meaning, data source, and summarizes data entity, data logic, and business source.

[0018] Preferably, the information extraction module integrates the business and data information by re-summarizing and refining the information at different levels in the corresponding business architecture (BA) and data architecture (IA), and records the association relationship between the business and the data.

[0019] Preferably, the knowledge generation module automatically generates declarative knowledge and procedural knowledge based on information extracted from the business architecture (BA) and the data architecture (IA), and enhances the knowledge to facilitate learning and use of large models, including:

[0020] Construct a function y=formalize{X} for knowledge standardization to ensure the consistency of knowledge, where X represents the original knowledge or data extracted from the business architecture (BA) and data architecture (IA), and y represents the output of the formalize function that conforms to the enterprise standard.

[0021] Preferably, the knowledge service module is responsible for storing and updating declarative knowledge and procedural knowledge, and providing knowledge services to the outside.

[0022] Preferably, the knowledge service module includes:

[0023] Knowledge creation: The business and data fusion knowledge documents automatically generated by the system are indexed through the large model LLM and stored in the vector library;

[0024] Knowledge query: Rewrite the question and complete the text sent by the user or system to the query interface according to the call parameters, and use the big model to form a summary of the conclusion and instruction actions;

[0025] Knowledge update: Real-time collection of updated content of business and data fusion knowledge documents, combined with feedback from knowledge query results, using large models to generate new summary knowledge, re-arrange indexes and update knowledge.

[0026] A method for constructing business and data fusion knowledge, applied to the system for constructing business and data fusion knowledge, comprises the following steps:

[0027] Step 1: Taking L1 (process classification) of the business architecture as an example, extract key business activities from L1 (process classification) of the business architecture, collect business and data information, and record the fusion model of business activities and data;

[0028] Step 2: Based on the business architecture (BA) and data architecture (IA), extract the data entities involved in business activities and the relationships between these data entities, collect the relationships between data entities and business activities and attach the upstream and downstream information of related business activities, and establish a fusion model of data entities and business;

[0029] Step 3: Based on the structure of declarative knowledge, automatically generate the declarative knowledge of business objects in the business architecture; automatically generate the declarative knowledge of data objects in the data architecture;

[0030] Step 4: Automatically generate procedural knowledge based on the structure of procedural knowledge. Procedural knowledge includes distributable execution, logical judgment, instruction mobilization, and data calculation.

[0031] Step 5: Add cross-level business and data relationships to knowledge and establish the level, object, and scope of knowledge applicability;

[0032] Step 6: Generate business and data fusion knowledge in a loop according to steps 1 to 5.

[0033] In summary, the present application includes at least one of the following beneficial effects:

[0034] 1. This application defines the framework of business and data fusion knowledge, forms the hierarchical structure, content composition, and expression method of business and data fusion knowledge, and establishes knowledge security management and authority management methods;

[0035] 2. Build a standard method for integrating business and data from the enterprise architecture (4A), extract business entities, business activities, business rules, business evaluation rules and other information from the business architecture (BA), and extract data entities, data attributes, data logic, data sources and other information from the data architecture (IA), automatically generate and enhance knowledge according to the knowledge framework, and improve the efficiency of knowledge construction;

[0036] 3. Use big models to establish methods for integrating business and data knowledge. When using data, you can clearly understand the relationship between data and business and what business problems the data can analyze and solve. Similarly, when analyzing business problems, the system can quickly locate the required data and provide analysis methods and results, thereby improving the efficiency of using data to solve business problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a schematic diagram of the overall system of Embodiment 1 of the present application;

[0038] Figure 2 It is a schematic diagram of the process of generating business and data fusion knowledge driven by the business architecture (BA) and data architecture (IA) in Example 2 of the present application. DETAILED DESCRIPTION

[0039] The following is in conjunction with the instruction manual Figure 1-2 This application is described in further detail.

[0040] Embodiment 1:

[0041] The embodiment of the present application discloses a system for constructing business and data fusion knowledge based on enterprise architecture, which is designed based on the methodology of enterprise architecture (4A); the enterprise architecture (4A) in this embodiment mainly includes business architecture (BA) and data architecture (IA). In addition, the system also includes a business and data fusion knowledge framework module, an information extraction module, a knowledge generation module, and a knowledge service module.

[0042] Among them, the business and data fusion knowledge framework module is responsible for defining the architecture and content of business and data fusion knowledge; it includes knowledge classification, knowledge classification, knowledge stratification, and knowledge content templates, specifically:

[0043] Knowledge classification designs a classification system for different knowledge, including the roles, permissions, encryption and desensitization requirements of knowledge, etc., laying a solid foundation for subsequent precise knowledge security management and control;

[0044] Knowledge classification and stratification combines the business concerns of L1 (process classification), L2 (process group), L3 (business process), L4 (industry activities), L5 (tasks) in the business architecture (BA), and uses the source information of business knowledge and data to correspond to the business attribution classification of knowledge. At the same time, the level of knowledge is recorded to facilitate the subsequent provision of more accurate knowledge results according to different types of roles;

[0045] The knowledge content template defines the type and structure of knowledge content, and provides a standard template for the subsequent automatic generation of business and data fusion knowledge based on the enterprise architecture (4A). This knowledge includes declarative knowledge and procedural knowledge. For example, an elephant is an animal, which is declarative knowledge; for example, putting an elephant in a refrigerator takes several steps, which is procedural knowledge. Declarative knowledge records comprehensive business definitions, instructions, usage methods, rules, etc.; procedural knowledge records knowledge such as processes, steps, methods, and actions.

[0046] Furthermore, the information extraction module is responsible for extracting the information required for generating knowledge from the business architecture BA and the data architecture IA, and increasing the relationship between business and data:

[0047] Collect basic information such as business segments, business processes, business activities, and business entities from L1 (process classification), L2 (process group), L3 (business process), L4 (industry activities), and L5 (tasks) of the business architecture (BA), and summarize business classification, business definition, business process, and business indicators;

[0048] Collect data entity, logical entity, attribute and other information from L1 (data domain), L2 (subject group), L3 (data entity), L4 (logical entity), and L5 (attribute) of the data architecture (IA), including data definition, data type, data meaning, data source and other information, and summarize data entity, data logic and business source;

[0049] The information summarized and refined at different levels such as L1, L2, L3, L4, and L5 in the business architecture (BA) and data architecture (IA) correspond one to one, integrating business and data information and recording the relationship between business and data.

[0050] Furthermore, the knowledge generation module is responsible for automatically generating declarative knowledge and procedural knowledge based on the information extracted from the business architecture (BA) and data architecture (IA), while enhancing the knowledge to facilitate learning and use of large models:

[0051] Construct a function y=formalize{X} for knowledge standardization to ensure knowledge consistency, where X represents the original knowledge or data extracted from the business architecture (BA) and data architecture (IA), and y represents the output processed by the formalize function into knowledge that meets enterprise standards;

[0052] According to the structure of declarative knowledge: formalize{[business level][business classification][business definition][business process][business indicator][data entity][related business][related relationship]}, the declarative knowledge of business objects such as business segments, business processes, business activities, business entities, business classification, business processes, and business indicators is automatically generated in the [L1 (data domain), L2 (subject group), L3 (data entity), L4 (logical entity), L5 (attribute)] level of the business architecture; the declarative knowledge of data objects such as data entities, logical entities, attributes, and indicator logic is automatically generated in the [L1 (data domain), L2 (subject group), L3 (data entity), L4 (logical entity), L5 (attribute)] level of the data architecture;

[0053] Based on the structure of procedural knowledge: formalize{[business rules][business judgment][process execution][instruction execution][indicator logic][data calculation][analysis method][applicable scenarios]}, automatically generate procedural knowledge such as distributable execution, logical judgment, instruction mobilization, data calculation, etc.;

[0054] Based on declarative knowledge and procedural knowledge, it automatically simulates manual Q&A to form a large model suitable for learning knowledge application scenarios.

[0055] Furthermore, the knowledge service module is responsible for storing and updating declarative knowledge and procedural knowledge, and providing knowledge services to the outside world. Specifically:

[0056] Knowledge creation: The business and data fusion knowledge documents automatically generated by the system are indexed through the large model LLM (also known as large language model) and stored in the vector library;

[0057] Knowledge query: Rewrite the question and complete the prompt words (i.e. a piece of text sent by the user or system to the query interface) according to the call parameters, and use the big model to form a summary of the conclusion and instruction actions;

[0058] Knowledge update: Real-time collection of updated content of business and data fusion knowledge documents, combined with feedback from knowledge query results, using large models to generate new summary knowledge, re-arrange indexes and update knowledge.

[0059] Embodiment 2:

[0060] The method for constructing business and data fusion knowledge based on enterprise architecture disclosed in the embodiment of the present application is applied to the system for constructing business and data fusion knowledge, and includes the following steps:

[0061] Step 1: Taking L1 (process classification) of the business architecture as an example, extract key business activities from L1 (process classification) of the business architecture; collect the definition of business activities, the business segment to which they belong, the required data, data indicators and other information, and record the fusion model of business activities and data, and call the formula formalize{Level,bizObj,dataObj,relation} to generate standardized terms.

[0062] Step 2: Based on the business architecture (BA) and data architecture (IA), extract the data entities involved in business activities and the relationships between these data entities; collect the relationships between data entities and business activities and attach the upstream and downstream information of related business activities, and establish a fusion model of data entities and businesses, and call the formula formalize{Level,bizObj,dataObj,relation} to generate standardized terms.

[0063] Step 3: Based on the structure of declarative knowledge: formalize{[business level][business classification][business definition][business process][business indicator][data entity][related business][related relationship]}, automatically generate the declarative knowledge of business objects such as business segments, business processes, business activities, business entities, business classifications, business processes, and business indicators in the [L1 (process classification), L2 (process group), L3 (business process), L4 (industry activity), L5 (task)] level of the business architecture (BA); automatically generate the declarative knowledge of data objects such as data entities, logical entities, attributes, and indicator logic in the [L1 (data domain), L2 (subject group), L3 (data entity), L4 (logical entity), L5 (attribute)] level of the data architecture (IA).

[0064] Step 4: Based on the structure of procedural knowledge: formalize{[business rules][business decisions][process execution][instruction execution][indicator logic][data calculation][analysis method][applicable scenarios]}, automatically generate procedural knowledge such as distributable execution, logical judgment, instruction mobilization, and data calculation.

[0065] Step 5: Add cross-level business and data relationships to the knowledge, and establish the level, object, and scope to which the knowledge applies.

[0066] Step 6: Follow the above five steps to cyclically generate business and data fusion knowledge at the L2, L3, L4, L5 and other levels.

[0067] Finally, in this embodiment 2, the core process of the business architecture (BA) and the data architecture (IA) driving the construction of business and data fusion knowledge is to extract the business information in the business architecture (BA) and the data information in the data architecture (IA), and then construct a complete knowledge framework of L1, L2, L3, L4, and L5. At the same time, by recording the generation, use, and update of business and data, the fusion relationship and information between the business and data are ensured to comprehensively form business and data fusion knowledge.

[0068] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A system for building business and data fusion knowledge based on enterprise architecture, including enterprise architecture (4A), characterized in that: It also includes business and data fusion knowledge framework module, information extraction module, knowledge generation module, and knowledge service module; The enterprise architecture (4A) includes business architecture (BA) and data architecture (IA); The business and data fusion knowledge framework module includes knowledge classification, knowledge classification, knowledge stratification, and knowledge content templates, wherein knowledge classification and knowledge stratification are respectively combined with the business concerns of the business architecture (BA), using the source information of business knowledge and data, corresponding to the business attribution classification of knowledge, and recording the level of knowledge; the knowledge content template is used to define the type and structure of knowledge content, and provide a standard template for the subsequent automatic generation of business and data fusion knowledge based on the enterprise architecture (4A); The information extraction module extracts information required for generating knowledge from the business architecture (BA) and the data architecture (IA), and increases the relationship between business and data; The knowledge generation module: extracts information from the business architecture (BA) and the data architecture (IA), and automatically generates declarative knowledge and procedural knowledge, while enhancing the knowledge; The knowledge service module stores and updates declarative knowledge and procedural knowledge, and provides knowledge services to the outside world.

2. A system for building business and data fusion knowledge based on enterprise architecture according to claim 1, characterized in that: The knowledge content template includes declarative knowledge and procedural knowledge. The declarative knowledge includes recording comprehensive business definitions, instructions, usage methods, and rules, and the procedural knowledge includes recording processes, steps, methods, and actions.

3. The system for building business and data fusion knowledge based on enterprise architecture according to claim 1, characterized in that: The business architecture (BA) includes L1 (process classification), L2 (process group), L3 (business process), L4 (industry activity), and L5 (task). The information extraction module collects business segments, business processes, business activities, and business entities, and summarizes them into business classification, business definition, business process, and business indicators.

4. A system for building business and data fusion knowledge based on enterprise architecture according to claim 3, characterized in that: The data architecture (IA) includes L1 (data domain), L2 (subject group), L3 (data entity), L4 (logical entity), and L5 (attribute). The information extraction module collects information such as data entity, logical entity, attribute, etc. from the data architecture (IA), including data definition, data type, data meaning, data source, and summarizes data entity, data logic, and business source.

5. The system for building business and data fusion knowledge based on enterprise architecture according to claim 1, characterized in that: The information extraction module integrates business and data information by summarizing and refining information at different levels in the corresponding business architecture (BA) and data architecture (IA), and records the association relationship between business and data.

6. The system for building business and data fusion knowledge based on enterprise architecture according to claim 1, characterized in that: The knowledge generation module automatically generates declarative knowledge and procedural knowledge based on information extracted from the business architecture (BA) and the data architecture (IA), and enhances the knowledge to facilitate learning and use of large models, including: Construct a function y=formalize{X} for knowledge standardization to ensure the consistency of knowledge, where X represents the original knowledge or data extracted from the business architecture (BA) and data architecture (IA), and y represents the output of the formalize function that conforms to the enterprise standard.

7. The system for building business and data fusion knowledge based on enterprise architecture according to claim 1 is characterized by: The knowledge service module is responsible for storing and updating declarative knowledge and procedural knowledge, and providing knowledge services to the outside world.

8. The system for building business and data fusion knowledge based on enterprise architecture according to claim 7, characterized in that: The knowledge service module includes: Knowledge creation: The business and data fusion knowledge documents automatically generated by the system are indexed through the large model LLM and stored in the vector library; Knowledge query: Rewrite the question and complete the text sent by the user or system to the query interface according to the call parameters, and use the big model to form a summary of the conclusion and instruction actions; Knowledge update: Real-time collection of updated content of business and data fusion knowledge documents, combined with feedback from knowledge query results, using large models to generate new summary knowledge, re-arrange indexes and update knowledge.

9. A method for constructing business and data fusion knowledge, applied to the system for constructing business and data fusion knowledge according to any one of claims 1 to 8, characterized in that: The following steps are involved: Step 1: Taking L1 (process classification) of the business architecture as an example, extract key business activities from L1 (process classification) of the business architecture, collect business and data information, and record the fusion model of business activities and data; Step 2: Based on the business architecture (BA) and data architecture (IA), extract the data entities involved in business activities and the relationships between these data entities, collect the relationships between data entities and business activities and attach the upstream and downstream information of related business activities, and establish a fusion model of data entities and business; Step 3: Based on the structure of declarative knowledge, automatically generate the declarative knowledge of business objects in the business architecture; automatically generate the declarative knowledge of data objects in the data architecture; Step 4: Automatically generate procedural knowledge based on the structure of procedural knowledge. Procedural knowledge includes distributable execution, logical judgment, instruction mobilization, and data calculation. Step 5: Add cross-level business and data relationships to knowledge and establish the level, object, and scope of knowledge applicability; Step 6: Generate business and data fusion knowledge in a loop according to steps 1 to 5.

Images