A maintenance updating method, device and equipment of an inventory service system

Abstract

The application relates to the field of computer software engineering and discloses a maintenance and updating method, device and equipment for an inventory business system, wherein the method comprises the following steps: acquiring heterogeneous original information of the inventory business system, generating target prompt text of the heterogeneous original information, calling a semantic extraction model based on the target prompt text to perform business semantic extraction, so as to obtain a target business field corresponding to a standard semantic segment, constructing an inventory business map according to the target business field, generating updating prompt information according to a business updating demand and the inventory business map in response to the business updating demand, performing business updating on the inventory business system according to the updating prompt information, and performing incremental updating on the inventory business map based on a business updating result. The technical scheme provided by the application can generate and maintain an inventory business map to synchronously update a system code library, and the efficiency and reliability of business inventory system maintenance and updating are improved.

Description

Technical Field

[0001] This application relates to the field of computer software engineering, and in particular to a method, apparatus and equipment for maintaining and updating an existing business system. Background Technology

[0002] Existing business systems, which originated in the early days of internet technology development, continue to serve as key carriers for the operation of core business operations. With the ongoing advancement of information technology infrastructure, intelligent maintenance and updates of these existing business systems are of paramount importance.

[0003] Due to limitations imposed by differing development standards, a lack of technical documentation, and personnel changes within the original development team, existing business systems generally exhibit characteristics such as information heterogeneity, complex calls, and missing logs. Traditional maintenance methods are inefficient and prone to overlooking critical dependencies, making maintenance quite difficult. Therefore, improving the efficiency and reliability of maintaining and updating existing business systems has become a key research focus in the field of existing business system maintenance. Summary of the Invention

[0004] This application provides a method, apparatus, and equipment for maintaining and updating existing business systems. It can generate and maintain existing business maps to synchronously update the system code library, thereby improving the efficiency and reliability of maintaining and updating existing business systems.

[0005] This application provides a method for maintaining and updating an existing business system. The method includes: acquiring multi-source heterogeneous information of the existing business system, wherein the multi-source heterogeneous information includes multiple heterogeneous raw information from different technology stacks; generating a standard semantic fragment of any of the heterogeneous raw information; extracting business semantics from the standard semantic field to obtain a target business field; constructing an existing business map based on the target business field; responding to a business update request initiated based on the existing business map; generating update prompt information based on the business update request and the existing business map; performing a business update on the existing business system based on the update prompt information; and incrementally updating the existing business map based on the business update result.

[0006] In one embodiment, generating a standard semantic segment of any of the heterogeneous original information includes: formatting the heterogeneous original information according to its information source to obtain initial semantic information; denoising the initial semantic information and segmenting the denoised initial semantic information to obtain the standard semantic segment of the heterogeneous original information.

[0007] In one implementation, extracting business semantics from the standard semantic fragment to obtain a target business field includes: generating target prompt text based on the standard semantic fragment and a preset prompt template; and inputting the target prompt text into a semantic extraction model to output the target business field of the standard semantic fragment through the semantic extraction model.

[0008] In one implementation, the target business field includes a business module field, a business process field, a function point field, and a data entity field; constructing an existing business map based on the target business field includes: for any target business field of the heterogeneous original information, creating corresponding hierarchical nodes for each field in the target business field; establishing multiple types of directed edges based on the node dependencies between the hierarchical nodes, the multiple types of directed edges including call edges, read / write edges, trigger edges, and association edges; and constructing the existing business map based on the hierarchical nodes and the multiple types of directed edges.

[0009] In one implementation, after constructing the existing business map based on the target business fields, the method further includes: rendering the existing business map in a specified form on the interactive interface of the existing business system, wherein the specified form is a tree diagram, flowchart, or topology diagram; performing a collaborative check on the existing business map on the interactive interface, updating the existing business map based on the collaborative check results, and storing the collaborative check results.

[0010] In one implementation, generating update prompt information based on the business update requirement and the existing business map includes: identifying target nodes related to the business update requirement in the existing business map, and identifying associated technical information of the target nodes; generating at least one clarification question based on a large language model according to the associated technical information and the business update requirement; and generating the update prompt information according to the feedback information for the clarification question and the associated technical information in response to the feedback information.

[0011] In one embodiment, updating the existing business system according to the update prompt information includes: sending the update prompt information to the target programming model or the target development end to instruct the target programming model or the target development end to update the code library of the existing business system according to the update prompt information, so as to complete the business update of the existing business system.

[0012] In one implementation, incrementally updating the existing business map based on the business update results includes: detecting change difference information in the codebase of the existing business system and determining the change type of the change difference information; determining the corresponding target update operation according to the change type, and determining the affected nodes and / or affected directed edges associated with the change difference information in the existing business map; and performing the target update operation on the affected nodes and / or affected directed edges to complete the incremental update of the existing business map.

[0013] In one embodiment, the method further includes: in response to a service deployment request initiated based on the existing business map, determining an initial node set related to the service deployment request; traversing multiple types of directed edges connecting any initial node in the initial node set to determine a target node set related to the service deployment request; extracting source data of any target node in the target node set, and generating a service deployment package based on the source data, wherein the service deployment package is used for independent deployment and operation in other system environments.

[0014] A second aspect of this application provides a maintenance and update apparatus for an existing business system. The apparatus includes: an information acquisition unit for acquiring multi-source heterogeneous information from the existing business system, wherein the multi-source heterogeneous information includes multiple heterogeneous raw information sources from different technology stacks; a map construction unit for generating a standard semantic fragment of any of the heterogeneous raw information sources, extracting business semantics from the standard semantic fragment to obtain a target business field, and constructing an existing business map based on the target business field; an update confirmation unit for responding to a business update request initiated based on the existing business map, generating update prompt information based on the business update request and the existing business map; and an update execution unit for performing a business update on the existing business system based on the update prompt information, and incrementally updating the existing business map based on the business update result.

[0015] A third aspect of this application provides a computer device, including: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the maintenance and update method for the existing business system as described in the first aspect above.

[0016] The technical solution provided in one or more embodiments of this application improves the efficiency and reliability of maintaining and updating existing business systems by generating and maintaining an existing business map to synchronously update the system codebase. Specifically, it determines standard semantic fragments based on the information source from multiple heterogeneous raw information from different technology stacks, uses a semantic extraction model to obtain target business fields by combining preset prompt templates, and then constructs an existing business map using the target business fields as hierarchical nodes. After the business update is completed, it performs incremental updates on affected nodes and directed edges by detecting changes in the codebase, thereby continuously and dynamically maintaining the business map and the codebase. This technical solution realizes unified semantic modeling and structured representation of heterogeneous information in an existing business system environment with multiple technology stacks coexisting. Based on the constructed existing business map, it performs interactive system code updates on demand and synchronously and dynamically updates the existing business map, thus improving the efficiency and reliability of maintaining and updating existing business systems. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the specific embodiments of this application or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 A schematic diagram illustrating the steps of a maintenance and update method for an existing business system provided in this application. Figure 2 This is a schematic diagram illustrating the steps of incremental map updating provided in one embodiment of this application; Figure 3 A schematic diagram illustrating the steps involved in generating a separate service deployment package as provided in one embodiment of this application; Figure 4 A schematic diagram illustrating the steps of a maintenance and update method for an existing business system provided in one embodiment of this application; Figure 5 A schematic diagram of the structure of a maintenance and update device for an existing business system provided in one embodiment of this application; Figure 6 This is a schematic diagram of the structure of a computer device provided in one embodiment of this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] Furthermore, the use of terms such as "first," "second," etc., in this application is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of embodiments in this application, unless otherwise stated, "multiple" means two or more. Additionally, the use of "based on" or "according to" implies openness and inclusiveness, because processes, steps, calculations, or other actions "based on" or "according to" one or more of the stated conditions or values ​​may in practice be based on additional conditions or beyond the stated values.

[0021] As enterprises continue to deepen their IT infrastructure development, a large number of existing business systems developed and put into operation in the early stages have become key carriers for core business operations. However, due to differences in development standards, lack of technical documentation, and personnel changes in the original development teams, these systems generally suffer from problems such as heterogeneous information from multiple sources, poor readability of business logic, blurred boundaries of functional modules, and difficulty in assessing the scope of impact of modifications, leading to a continuous increase in maintenance and iteration costs in the later stages.

[0022] Specifically, existing systems typically contain multiple technical implementations from various sources, including backend business code, database schema definitions, and frontend page components, resulting in a highly heterogeneous technology stack. For example, within the same enterprise, there may be system instances built on multiple technology systems, such as J2EE (Java 2 Platform Enterprise Edition, a Java development platform), .NET Framework (a managed execution environment for Windows operating systems), PHP (Hypertext Preprocessor, an open-source server-side scripting language), and the earlier Spring (an open-source J2EE application framework). Existing code analysis tools can only process information from a single source, leading to a reliance on manual piecing together of information for a comprehensive understanding of the business, which is inefficient and prone to overlooking critical dependencies. Furthermore, after years of evolution and iteration, the inter-module call relationships are complex, and the code coupling is high, making it extremely easy for manual analysis to miss critical dependencies, leading to runtime anomalies or data inconsistencies after subsequent splitting and deployment.

[0023] In related technologies, code interpretation schemes based on large language models typically generate one-time natural language documents or diagrams. However, this approach cannot detect subsequent additions or deletions of functions and code changes to the system. As the existing system undergoes continuous iteration, discrepancies arise between the static documentation and the actual code implementation, leading to a decrease in documentation credibility and an inability to support accurate maintenance decisions. Furthermore, existing microservice decomposition or low-code deployment solutions are mostly geared towards newly built systems or rely on standardized component libraries, making it difficult to directly apply to existing, tightly coupled codebases.

[0024] In view of this, one or more embodiments of this application provide a method, apparatus and system for maintaining and updating an existing business system, which performs interactive system code updates on demand based on a constructed existing business map, and synchronously and dynamically updates the existing business map, so as to improve the efficiency and reliability of maintaining and updating the existing business system.

[0025] Please see Figure 1 One embodiment of this application provides a method for maintaining and updating an existing business system, which may include the following steps: S1: Obtain multi-source heterogeneous information from existing business systems, wherein the multi-source heterogeneous information includes multiple heterogeneous original information from different technology stacks; The aforementioned multi-source heterogeneous information can be understood as multiple heterogeneous raw information sources originating from different technology stacks within the existing business system. This heterogeneous raw information represents unprocessed text information collected within the corresponding technology stack. For existing business systems, their codebases may contain multiple technology stacks, each with different file syntax structures, parsing methods, and code types. For example, backend code often uses Java source files, databases often use SQL statements, and frontend pages often use Vue component code.

[0026] Generally, existing business systems comprise three independent technology stacks: backend code, database, and frontend pages. The heterogeneous raw information that needs to be collected differs for each stack. Specifically, for the backend code technology stack, the heterogeneous raw information can include fragments such as interface definitions, function call chains, SQL statements, class / method / package structures, and comments. For the database technology stack, the heterogeneous raw information can include fragments such as table structures, constraints, indexes, triggers, stored procedures, sample data, and comments. For the frontend page technology stack, the heterogeneous raw information includes fragments such as component trees, route mappings, API call mappings, and user interaction events.

[0027] The aforementioned heterogeneous raw information is collected using different methods depending on the technology stack. Specifically, heterogeneous raw information from the backend code is obtained by parsing the codebase using an Abstract Syntax Tree (AST). Heterogeneous raw information from the database is collected by connecting to the database instance. Heterogeneous raw information from the frontend page is obtained by parsing the frontend project files. After the three types of information are collected simultaneously, they are aggregated into multi-source heterogeneous information.

[0028] S3: Generate a standard semantic fragment of any of the heterogeneous original information, extract business semantics from the standard semantic fields to obtain target business fields, and construct an existing business map based on the target business fields; In this embodiment, the heterogeneous raw information is formatted, denoised, and unified in data structure to obtain corresponding standard semantic fragments. Each standard semantic fragment is a JSON object. These standard semantic fragments can be categorized according to fragment type. For example, taking heterogeneous raw information of the backend code class as an example, its standard semantic fragments may include interface definition fragments, function call chain fragments, and SQL statement fragments. For example, an interface definition fragment can be represented as: '{"path":" / api / patient / register","method":"POST","parameters":[...],"response":...}'.

[0029] In this embodiment, the aforementioned business semantic extraction can be understood as extracting key business fields from standard semantic fields as target business fields. Specifically, business semantic extraction can be performed on standard semantic fields using a large language model. Optionally, the standard semantic fields can be directly input into the large language model. Optionally, standard semantic fragments can be integrated and extracted based on a preset prompt template, and the resulting prompt text can be input into the large language model to standardize the input and obtain the target business fields more accurately and quickly. Optionally, the aforementioned target business fields are represented in the form of business quadruples, specifically including field information such as business modules, business processes, functional points, and data entities.

[0030] In this embodiment, a business inventory map is constructed based on the field information and the dependencies represented in the target business fields. This business inventory map can be understood as a directed, hierarchical knowledge graph. Specifically, for each target business field, nodes are constructed by dividing the field information into four levels, and all target business fields under multi-source heterogeneous information are constructed into nodes according to the above hierarchical form. Further, directed edges are constructed between corresponding nodes based on the dependencies between the field information. Specifically, these dependencies can be divided into call relationships (e.g., function A calls function B, determined based on the call chain), read-write relationships (e.g., read or write operations of function points on data entities, determined based on SQL statements), trigger relationships (e.g., front-end events trigger back-end APIs, determined based on page API mapping), and association relationships (e.g., database foreign key constraints or natural business associations, determined based on database schema).

[0031] Optionally, to facilitate subsequent persistent storage, transaction management, and efficient querying, the existing business map can be recorded in the form of a relational table structure. Specifically, the relational table structure is built hierarchically based on the existing business map and can include business module tables, business process tables, function point tables, data entity tables, and dependency relationship tables. This relational table structure storage method reduces the existing business system's dependence on graph databases, facilitating deployment and maintenance.

[0032] S5: In response to a business update request initiated based on the existing business map, generate an update prompt message according to the business update request and the existing business map; The aforementioned business update requirements refer to descriptions of modifications or additions to a specific functional module of an existing business system, initiated by a user or the system. The update prompt information is a structured modification instruction generated based on the context of the existing business map and the business update requirements. It is used to transform business requirements in natural language form into precise technical descriptions that can directly guide code modification. Specifically, the project maintainer initiates a business update request in the existing business system. The business update request points to a specific node or set of nodes in the existing business map, and carries the business update requirements.

[0033] Since business update requirements are typically expressed in natural language, directly sending the requirement description to the programming model or development team can easily lead to inaccurate or completely wrong changes due to misunderstandings of the requirements. Optionally, by first locating target nodes and extracting related technical information in the existing business map, and then using a large language model to generate update prompts, the system can address this issue. For example, update prompts can include structured fields such as target node identifiers, descriptions of the change scope, related interface definitions, data model change descriptions, and expected impact assessments. By automatically converting vague business requirements into precisely located, structured change prompts based on the existing business map, the risk of misunderstandings of change requirements is significantly reduced, improving the accuracy and controllability of subsequent code updates.

[0034] S9: Based on the update prompt information, perform business updates on the existing business system, and perform incremental updates on the existing business map based on the business update results.

[0035] The aforementioned business update operation refers to the process of modifying the codebase of an existing business system based on update prompts. The executing entity can be the target programming model or the target development end (human developers). The target programming model can utilize an automated programming assistant (such as the Copilot programming tool). The aforementioned incremental update operation refers to performing partial updates on affected nodes and directed edges in the existing business map based on the changes in the codebase after the business update is completed, to maintain continuous dynamic consistency between the business map and the codebase state.

[0036] Compared to traditional technologies that generate one-time static business documents or diagrams based on large language models, this implementation minimizes the maintenance cost of the business map by automatically detecting codebase changes and performing local incremental updates after each business update, while ensuring it always reflects the latest state of the codebase. The incremental update mechanism achieves continuous dynamic synchronization between the business map and the codebase state, eliminating the problems of static documents becoming invalid with system iterations and failing to detect subsequent code changes, leading to discrepancies between the document and the code implementation. When subsequent updates and maintenance are needed, the maintained existing business map can be directly invoked for interactive updates, improving the maintainability of the existing business system and thus enhancing the efficiency and reliability of maintaining and updating the existing business system.

[0037] Based on the above ideas, the technical solution provided in this embodiment of the application improves the efficiency and reliability of maintaining and updating existing business systems by generating and maintaining an existing business map to synchronously update the system codebase. Specifically, standard semantic fragments are determined based on the information source from multiple heterogeneous raw information from different technology stacks. A semantic extraction model is then used in conjunction with a preset prompt template to obtain target business fields. These target business fields are then used as hierarchical nodes to construct an existing business map. After a business update is completed, incremental updates are performed on affected nodes and directed edges by detecting changes in the codebase, thus enabling continuous dynamic maintenance of the business map and the codebase. This technical solution achieves unified semantic modeling and structured representation of heterogeneous information in an existing business system environment with multiple technology stacks. Based on the constructed existing business map, interactive system code updates are performed on demand, and the existing business map is updated synchronously and dynamically, improving the efficiency and reliability of maintaining and updating existing business systems.

[0038] In one implementation, because the raw information from different technology stacks differs fundamentally in syntactic structure and expression, and excessively long contexts can easily lead to semantic confusion and slow information processing efficiency, it is necessary to transform heterogeneous raw information from different technology stacks into standardized semantic fragments with unified structure and reduced noise to further improve the accuracy and stability of business semantic extraction. Specifically, the following steps are followed to generate a standardized semantic fragment for any heterogeneous raw information: S21: Based on the information source of the heterogeneous original information, format the heterogeneous original information to obtain the initial semantic information; S23: Denoise the initial semantic information and segment the denoised initial semantic information to obtain standard semantic segments of the heterogeneous original information.

[0039] Formatting refers to the structural transformation of heterogeneous original information based on its source (i.e., corresponding technology stack) using differentiated processing strategies to obtain initial semantic information. Specifically, according to the syntax rules and data structure characteristics corresponding to different information sources, appropriate parsing tools are used to uniformly convert the information into JSON format. Converting to JSON format facilitates model understanding. Optionally, the heterogeneous original information can be denoised first before format conversion.

[0040] The noise reduction process involves filtering out redundant content (such as comments, blank lines, and test code) unrelated to business semantics from the initial semantic information, retaining only the valid content with business semantic value. Specifically, filtering is performed on the initial semantic information to remove code comments, blank lines, unit test code, debug logs, and other content irrelevant to business semantics, while retaining valid content such as interface definitions, business logic functions, and data model definitions. This valid content is then segmented according to fragment type to determine standard semantic fragments. For example, standard semantic fragments can be determined by specifying keywords. For any fragment type, there is a corresponding specified keyword, and the code region (function-level, table-level, or page-level) containing the specified keyword is used as the standard semantic fragment.

[0041] The technical solution provided in this embodiment refines the processing method for converting heterogeneous raw information into unified standard semantic fragments. Specifically, backend code, database schema, and frontend page heterogeneous raw information are converted into JSON format, eliminating format barriers caused by differences in technology stacks. Noise reduction filtering filters out invalid content such as comments, blank lines, and test code, reducing input noise for subsequent models and improving the signal-to-noise ratio of semantic understanding. Furthermore, the formatted and noise-reduced heterogeneous raw information is segmented according to fragment type, ensuring that each standard semantic fragment is semantically focused and has clear boundaries, avoiding semantic confusion caused by long contexts. Therefore, this technical solution standardizes the preprocessing flow of multi-source heterogeneous information, effectively improving the accuracy of the target business fields output by the subsequent semantic extraction model, thereby improving the efficiency and reliability of maintaining and updating existing business systems.

[0042] In one implementation, extracting business semantics from standard semantic fragments to obtain target business fields includes the following steps: S31: Generate target prompt text based on standard semantic fragments and preset prompt templates; S33: Input the target prompt text into the semantic extraction model to output the target business field of the standard semantic fragment through the semantic extraction model.

[0043] The aforementioned preset prompt template includes role settings, context descriptions, specific content embedding, and output format constraints. The context descriptions inform the semantic extraction model which type of standard semantic fragment is being analyzed. The specific content embedding represents key information within the standard semantic fragment. The output format constraints instruct the semantic extraction model to output according to the target business fields. These target business fields include business module fields, business process fields, function point fields, and data entity fields. The semantic extraction model employs a large language model.

[0044] In one embodiment, for standard semantic fragments defined by the backend API interface, the target prompt text for inputting into the large language model is generated according to a preset prompt template as follows: Role setting: You are a senior business systems analysis expert; Contextual explanation: Analyze the following backend API interface definitions to identify the business modules, business processes, functionalities, and related data entities; Embedded details: Path: {'api_path'}; Method: {'http_method'}; Parameters: {'parameters'}; Response: {'response'}; Related SQL: {'sql_statements'}; Class: {'class_name'}; Method: {'method_name'}; Output format constraints: Please strictly follow the following JSON format for output, and do not include any additional explanations: {"business_module":"Business module name: Appointment registration module","business_process":"Business process name: Online registration process","function_point":"Function point description: Create registration","data_entities":"Data entities involved: registration table"}.

[0045] In this embodiment, semantic extraction is performed on the target prompt text based on a large language model to obtain the target business fields, which may be: "Business Module: 'Appointment Registration Module'; Business Process: 'Online Registration Process'; Function Point: 'Create Registration'; Data Entity: 'Registration Table'."

[0046] In this embodiment, the target business fields include business module fields, business process fields, function point fields, and data entity fields. Constructing an existing business map based on the target business fields includes the following steps: S41: For any target business field of heterogeneous original information, create corresponding hierarchical nodes for each field in the target business field; S43: Establish multiple types of directed edges based on the node dependencies between hierarchical nodes. These multiple types of directed edges include call edges, read / write edges, trigger edges, and association edges. S45: Build a map of existing business based on hierarchical nodes and multiple types of directed edges.

[0047] Specifically, an empty graph is created, and the existing business map is constructed by traversing the representation fields and node dependencies of each target business field. The hierarchical nodes include first-level nodes (representing business module fields), second-level nodes (representing business process fields), third-level nodes (representing function point fields), and leaf nodes (representing data entity fields). Multiple types of directed edges represent function point calls, function point read / write operations on data entities, front-end event triggering back-end APIs, and database foreign key associations, respectively. Optionally, node attributes such as name, description, source code location, and last update timestamp are stored.

[0048] The technical solution provided in this embodiment transforms target business fields into a hierarchical and computable existing business map through semantic annotation and structured construction. Specifically, it calls a semantic extraction model to annotate standard semantic fragments with business semantics, outputting target business fields containing business module fields, business process fields, function point fields, and data entity fields, thus achieving automatic mapping from technical code to business semantics. By creating corresponding hierarchical nodes for each field and establishing four types of directed edges—call edges, read-write edges, trigger edges, and association edges—based on node dependencies, it endows the business map with computable graph structure characteristics. Furthermore, the existing business map can be converted into a relational data structure for persistent storage, ensuring that the business map supports efficient node query and incremental update operations, providing a visual foundation for subsequent accurate updates and service reorganization. Simultaneously, the automated semantic annotation and graph construction process allows for the rapid construction of the entire system's business map, improving the efficiency of maintaining and updating the existing business operating system.

[0049] In one implementation, in existing business systems with partially ambiguous boundaries or non-standard naming, to avoid issues such as incorrect node attribution or missing dependencies, a visual collaborative inspection interface is provided. This allows project maintenance personnel with domain knowledge to manually verify and correct the automatically constructed results, improving the accuracy of the existing business map. Specifically, after constructing the existing business map based on the target business fields, the method further includes the following steps: S51: Render the existing business map in a specified form on the interactive interface of the existing business system, wherein the specified form is a tree diagram, flowchart or topology diagram. S53: Perform collaborative checks on the existing business map on the interactive interface, update the existing business map based on the collaborative check results, and store the collaborative check results.

[0050] In this embodiment, the existing business map is presented as a tree diagram, flowchart, or topology. Figure 3One form of visualization is presented on the interactive interface of the existing business system, enabling project maintenance personnel to intuitively perceive the hierarchical structure and dependencies of the existing business map. Based on this, project maintenance personnel collaboratively check the automatically constructed existing business map on the interactive interface, manually verifying each node and various types of directed edges. For example, error type checks are performed according to Table 1 below to complete the collaborative check of the existing business map. Corrections are then made according to the corresponding error types to update the existing business map.

[0051] Table 1 Error types and corresponding correction methods The aforementioned collaborative inspection results characterize the updated existing business map and modification record information, which may include, for example, node metadata, edge metadata, target business fields, manual correction records, and map-level metadata. For instance, the node metadata may include node ID, node name, business type, business description, source location, confidence level, creation method, timestamp, modifier, and version number; the edge metadata may include edge ID, source / target node, dependency type, business semantics, source basis, confidence level, creation method, and manual overwrite marker; the manual correction records include operation type, before and after snapshots, operator, and timestamp; and the map-level metadata may include map ID, system name, version, status, and statistical information.

[0052] The technical solution provided in this embodiment renders the existing business map in a specified format on an interactive interface, and supports project maintenance personnel to perform collaborative inspection operations such as node renaming, hierarchy adjustment, type modification, connection addition and deletion, and node addition and deletion on the interactive interface, realizing a human-computer collaborative confirmation mechanism for automatically constructing the business map. Specifically, the visual rendering method allows project maintenance personnel to intuitively perceive the hierarchical structure and dependencies of the business map, and the collaborative inspection method, by providing various editing operations, allows maintenance personnel with domain knowledge to accurately correct the automatically constructed results. This technical solution further improves the accuracy and traceability of the business map, reducing the risk of misoperation due to map errors in subsequent maintenance decisions.

[0053] In one implementation, update prompt information is generated based on a large language model according to business update requirements and an existing business map, specifically including the following steps: S61: In the existing business map, identify the target nodes related to business update requirements and determine the associated technical information of the target nodes; S63: Based on the relevant technical information and business update requirements, generate at least one clarification question based on the large language model; S65: In response to feedback regarding clarification issues, generate an updated prompt message based on the feedback message and related technical information.

[0054] The aforementioned target nodes are hierarchical nodes in the existing business map that are directly related to business update requirements. Within the existing business map, target nodes are determined by locating nodes based on the semantic content of the business update requirements, or directly in response to user interactions with nodes on the human-computer interface. The aforementioned associated technical information is the technical context associated with the target node in the business map, including interface definitions, data models, and call chain information. This information can be determined based on the associated nodes and edge information connected to the target node.

[0055] In this implementation, a joint analysis of related technical information and business update requirements is performed based on a large language model. At least one clarification question is generated and user responses are received to eliminate ambiguity before an update prompt is generated. When multiple clarification questions are generated, a structured list of clarification questions is created and user responses are received. Examples of possible responses include: whether a new data table needs to be created for the new feature, whether existing interface signatures need to be modified, and whether access control is involved. After receiving supplementary confirmation from the user, a structured update prompt is generated based on the confirmation result and related technical information. This update prompt may include structured fields such as the modification target, involved code location, specific modification requirements, precautions, and reference code examples to indicate the update content for the existing business system.

[0056] The technical solution provided in this embodiment generates structured update prompts by locating target nodes and extracting associated technical information in the existing business map, and supplementing and confirming this information using a large language model. Specifically, by determining target nodes and their associated technical information in the existing business map, and supplementing and confirming this information using a large language model to generate structured update prompts, the system can accurately identify the targets for modification based on fuzzy business needs. This effectively reduces the risk of misunderstanding modification requirements and improves the accuracy and controllability of maintaining and updating existing business systems.

[0057] In one implementation, a target programming model or a target development end is used to update the existing business system. Optionally, update prompts are sent to the target programming model to instruct it to update the codebase of the existing business system according to the update prompts, thereby completing the business update. Optionally, update prompts are also sent to the target development end to instruct it to update the codebase of the existing business system according to the update prompts, thereby completing the business update.

[0058] Specifically, the target programming model can be an AI-powered programming model with code generation capabilities, capable of automatically generating relevant code conforming to the target language specifications based on structured update prompts to perform code updates. For example, the Copilot programming tool can be used as the target programming model. The target development end can be a workstation where human developers perform code updates, such as an integrated development environment, code review system, or task management platform. The code update operation specifically involves adding, modifying, or deleting relevant code in the existing business system's codebase based on update prompts to achieve business updates for corresponding functions.

[0059] The technical solution provided in this embodiment supports two business update execution paths: automated code generation and manual code modification, by sending structured update prompts to the target programming model or the target development end. The target programming model execution path automatically generates corresponding code changes and submits them to the code repository based on the modification scope, interface definitions, and data model change descriptions in the update prompts, making it suitable for scenarios with high standardization and low modification risk. The target development end execution path pushes the update prompts to developers in structured form, allowing them to modify the code according to the prompts and submit it to the code repository, making it suitable for high-security business scenarios. This technical solution, while ensuring the accuracy of modification instructions, flexibly adapts to the maintenance and update scenarios of existing business systems with different security levels and automation requirements, improving the practicality and applicability of the solution.

[0060] In one implementation, please refer to Figure 2 Incremental updates to existing business maps based on business update results include: S81: Detect change difference information in the codebase of existing business systems and determine the change type of the change difference information; S83: Based on the change type, determine the corresponding target update operation, and in the existing business map, determine the affected nodes and / or affected directed edges associated with the change difference information; S85: Perform target update operations on affected nodes and / or affected directed edges to complete incremental updates to the existing business map.

[0061] The aforementioned change difference information refers to the differences in code content before and after a business update in the codebase, typically representing three types of change operations: addition, modification, and deletion. The aforementioned change type is a semantic classification of the change difference information, used to determine the corresponding target update operation (e.g., node addition, node modification, node deletion, edge addition, edge deletion, etc.). The aforementioned affected nodes are hierarchical nodes in the existing business map that have a semantic relationship with the change difference information, and the affected directed edges are directed edges in the existing business map that have a semantic relationship with the change difference information. The target update operation is the specific update action performed on the affected nodes and / or affected directed edges, determined based on the change type.

[0062] Specifically, the system detects change discrepancies in the codebase of existing business systems. By comparing the code content before and after the update, it extracts three types of change discrepancies: additions, modifications, and deletions, and determines the change type for each. Based on the change type, the corresponding target update operation is determined: for example, for interface addition changes, the target update operation is to add the corresponding functional point node and its associated call edge in the business map; for data model modification changes, the target update operation is to update the attributes of the corresponding data entity node; for interface deletion changes, the target update operation is to delete the corresponding functional point node and its associated directed edge. In the existing business map, the affected nodes and / or affected directed edges associated with the change discrepancies are identified, and the target update operation is performed on the affected nodes and / or affected directed edges to complete the incremental update of the existing business map.

[0063] Table 2 A set of possible incremental update operations In one embodiment, the code content before and after the codebase update is compared to determine the change differences, for example, by using the `git diff` command to view code differences. Referring to Table 2 above, the change differences are categorized into corresponding change types, mapped to the corresponding elements in the existing business map (affecting nodes or directed edges), and corresponding operations are performed. For each type of change, the system performs the corresponding update operation to avoid a full map reconstruction. Optionally, after the automatic update is complete, a consistency check can be performed on the map, such as: dangling edge check (edges pointing to deleted nodes), isolated node check (no edges connecting to other nodes), and high-risk change check (such as deleting a feature point that is depended on by multiple other nodes).

[0064] The specific operations for adding nodes are as follows: Newly identified API / table creation function points or data entity nodes are selected, with the creation method set to AUTO_UPDATE and a medium confidence level, pending user confirmation. Modifying node attributes involves directly updating the node's JSON fields (e.g., api_signature) and incrementing the version number, recording the last modification time. Deleting nodes / edges involves not immediately deleting nodes from the map for elements already deleted in the code; instead, marking them as obsolete and prompting the user for confirmation. If the user confirms, the node is removed; otherwise, it is retained for historical reference. Adding new edges involves creating new directed edges between relevant nodes based on new call relationships, front-end API mappings, etc., and annotating the dependency type and source basis.

[0065] The technical solution provided in this embodiment refines the incremental update mechanism for existing business maps. Specifically, by comparing the content before and after the code repository update, it extracts three types of change difference information: addition, modification, and deletion. This change difference information is then mapped to target update operations such as node addition, node modification, node deletion, edge addition, and edge deletion, ensuring a precise correspondence between update operations and change semantics. Simultaneously, based on the change difference information, the affected local areas of the existing business map are located, and target update operations are performed only on the affected nodes and directed edges, preserving the historical state of unaffected parts and avoiding the high cost of full reconstruction. Therefore, this technical solution achieves continuous dynamic synchronization between the business map and the code repository state, eliminating the problem of static documents becoming invalid with system iterations in traditional technologies. It ensures that each maintenance decision is based on an existing business map consistent with the code repository state, while minimizing the computational overhead of business map maintenance.

[0066] In one implementation, please refer to Figure 3 The method can also generate service deployment packages based on existing business maps, which can then run independently on other systems to achieve corresponding functions. Specifically, follow these steps: S91: In response to a service deployment request initiated based on an existing business map, determine the initial set of nodes related to the service deployment request; S93: Traverse the multi-type directed edges connecting any initial node in the initial node set to determine the target node set that meets the service deployment requirements; S95: Extract source data from any target node in the target node set, generate a service deployment package based on the source data, and use the service deployment package for independent deployment and operation in other system environments.

[0067] In step S91 above, the service deployment requirement can be understood as a user's request to extract a functional module from the existing business system and deploy and run it independently in another system environment. Optionally, the service deployment requirement can be represented by the user initiating a service deployment request on the human-computer interaction interface of the existing business system. This request is confirmed by the user selecting one or more functional nodes using checkboxes on the business map visualization interface (tree diagram or topology diagram). The selected objects can be business modules, business processes, or functional points, typically with functional points as the smallest granularity. For example, when a user selects a business module, the system defaults to selecting all its sub-nodes (business processes and functional points), and the user can manually deselect individual sub-nodes.

[0068] In step S93 above, the directed edges connecting any initial node in the initial node set are traversed, that is, along the call edge, read / write edge, trigger edge, and association edge, all nodes connected to that initial node are traversed. The initial node set is expanded by performing a breadth-first traversal to obtain the target node set, which contains each initial node and all its connected nodes.

[0069] In step S95 above, the source data includes corresponding source code files or function blocks, database table definitions, configuration files, and a list of third-party library dependencies. This source data can be determined based on the source identifiers recorded at each node in the existing business map. The source identifiers can represent file paths, class names, method names, and SQL table names. The relevant content under the source identifiers is then extracted from the original code repository and database as a service deployment package. For example, the service deployment package may include a base image and build instructions, a service orchestration configuration file, an environment variable template file, a list of third-party library dependencies, and a database incremental trimming script.

[0070] The technical solution provided in this embodiment offers a method for generating service deployment packages based on an existing business map. Specifically, based on service deployment requirements, the starting node is accurately located in the business map. Dependency nodes are expanded through breadth-first traversal, and the transitive closure of the target node set is automatically calculated to ensure that all necessary dependencies are included. Furthermore, the source code files, DDL statements, and environment configuration items corresponding to each target node are precisely located from the code repository to create the service deployment package. This technical solution effectively solves the dependency omission and redundancy problems in the modular deployment of existing monolithic systems, improving the efficiency and reliability of service-oriented transformation of existing business systems.

[0071] Please see Figure 4This application provides one or more embodiments of the above-described method for maintaining and updating an existing business system. This embodiment uses a 15-year-old hospital information system of a tertiary hospital as an example scenario, aiming to modify and split the "appointment registration" related functions into independent private services. The specific steps are as follows: S101: Respond to system maintenance operation commands and obtain multi-source heterogeneous information from existing business systems.

[0072] Specifically, maintenance personnel trigger a multi-source information collection operation on the interactive interface, specifying the code repository address, database connection information, and front-end project path of the existing system. The system responds to these instructions, collecting multi-source heterogeneous information in parallel via three channels. Based on the information source of each type of heterogeneous raw information, it performs differentiated formatting processing, uniformly converting the backend interface definitions into JSON format and generating a corresponding set of standard semantic fragments.

[0073] S102: Respond to the business map construction operation command and construct the existing business map.

[0074] Specifically, after the collection of multi-source heterogeneous information is completed, the system responds to the map construction operation command, inputs the standard semantic fragment set into the semantic extraction model, performs business semantic annotation on each fragment, and outputs target business fields in JSON format. These target business fields include business module fields, business process fields, function point fields, and data entity fields. The semantic extraction model identifies business module fields such as registration management, patient management, and scheduling management; business process fields such as appointment registration, cancellation registration, and payment confirmation; function point fields such as appointment source query, status update, and payment callback; and data entity fields such as the registration record table, patient table, scheduling table, and appointment source table.

[0075] Furthermore, create corresponding hierarchical nodes for each field in the target business field. The business modules corresponding to fields such as registration management, patient management, and scheduling management are designated as first-level nodes; the business processes corresponding to fields such as appointment registration and cancellation registration are designated as second-level nodes; the functional points corresponding to fields such as appointment source query and status update are designated as third-level nodes; and the data entities corresponding to fields such as registration record table and patient table are designated as leaf nodes. Based on the node dependencies between hierarchical nodes, establish multiple types of directed edges: call edges based on backend interface call chains, read / write edges based on database read / write operations, trigger edges based on payment callback event triggering relationships, and association edges based on foreign key constraints. Based on the hierarchical nodes and multiple types of directed edges, build an existing business map.

[0076] S103: Responds to map collaboration confirmation operation commands and performs visualization rendering and collaborative checks on existing business maps.

[0077] Specifically, the existing business map is rendered as a topology graph on the interactive interface, displaying the network of all node dependencies and the call edges, read / write edges, trigger edges, and association edges between nodes. Senior hospital technicians collaboratively inspect the automatically constructed business map on the interactive interface and discovered that the semantic extraction model had missed the "blacklist verification" function node (this existing system checks whether patients are blacklisted before appointment registration). They added the "blacklist verification" function node under the "appointment registration" business process node via the right-click menu and dragged and dropped to add the call edge between it and the "appointment registration" function node. The existing business map is updated based on the collaborative inspection results, and the results are stored. The stored content includes the operation type (node ​​addition, edge addition), snapshots of the node / edge before and after the operation, the operator's identifier, and the operation timestamp, supporting version snapshot recording and backtracking.

[0078] S104: Respond to business update request input and generate update prompt information.

[0079] Specifically, maintenance personnel select the "Appointment Booking" business process node on the interactive interface and input the business update requirement in natural language: "In the appointment booking process, after a patient successfully pays via WeChat, the system should automatically send a confirmation message and update the booking status to confirmed; at the same time, if a doctor temporarily cancels their appointment, a notification should be automatically pushed and the patient should be allowed to change their appointment with one click." The system responds to the above business update requirement by identifying the target nodes related to the requirement in the existing business map as the "Payment Callback" and "Status Update" function nodes under the "Appointment Booking" business process node, and determining the associated technical information of the target nodes. Based on the associated technical information and the business update requirement, supplementary confirmation is performed using a large language model. The large language model generates a structured list of clarification questions. After receiving the supplementary confirmation information from the maintenance personnel, a structured update prompt message is generated based on the supplementary confirmation results and the associated technical information.

[0080] S105: Respond to business update operations, perform business updates on existing business systems, and perform incremental updates on existing business maps.

[0081] Specifically, update prompts are sent to the target programming model. Based on the modification scope description, associated interface definitions, and data model change instructions in the update prompts, the target programming model automatically generates corresponding code changes and submits them to the system code repository. After the code changes are submitted, the system detects the change difference information in the code repository. By comparing the differences before and after the code repository update, the system extracts the change difference information, determines the corresponding target update operation based on the change type, and identifies the affected nodes in the existing business map. The system then performs the target update operation on the affected nodes and their directed edges. For example, under the "Appointment Registration" business process node, a "Send Confirmation Message" function node is added, the attributes of the "Registration Record" data entity node are updated (a new status enumeration value is added), and a trigger edge is added between the "Payment Callback" function node and the "Send Confirmation Message" function node. This completes the incremental update of the existing business map and achieves continuous dynamic synchronization between the business map and the code repository status.

[0082] S106: Respond to service deployment requirements and generate a service deployment package.

[0083] Specifically, hospital operations and maintenance personnel select the "Appointment Booking" service module node on the interactive interface, triggering a service deployment request. The system responds to this service deployment request by determining an initial node set, which includes the "Appointment Booking" service module node (a first-level node). It then traverses the directed edges connecting the "Appointment Booking" service module node in the initial node set to automatically determine the target node set, which includes the appointment booking node and all its transitive dependencies (patient information query node, blacklist verification node, and related database table nodes). The system extracts the source data for each target node in the target node set and generates a service deployment package based on this data. This service deployment package is used for independent deployment and operation in the branch hospital server environment.

[0084] Please see Figure 5 This application also provides a maintenance and update device for an existing business system, the device comprising: The information acquisition unit 100 is used to acquire multi-source heterogeneous information from existing business systems, wherein the multi-source heterogeneous information includes multiple heterogeneous original information from different technology stacks. The map construction unit 200 is used to generate a standard semantic fragment of any of the heterogeneous original information, extract business semantics from the standard semantic fields to obtain target business fields, and construct an existing business map based on the target business fields. The update confirmation unit 300 is used to respond to a business update request initiated based on the existing business map, and generate update prompt information according to the business update request and the existing business map; The update execution unit 400 is used to perform business updates on the existing business system according to the update prompt information, and to perform incremental updates on the existing business map based on the business update results.

[0085] In one embodiment, the apparatus further includes: a service configuration unit 500, configured to, in response to a service deployment request initiated based on the existing service map, determine an initial node set related to the service deployment request; traverse multiple types of directed edges connecting any initial node in the initial node set to determine a target node set related to the service deployment request; extract source data of any target node in the target node set; and generate a service deployment package based on the source data, wherein the service deployment package is used for independent deployment and operation in other system environments.

[0086] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0087] In this application embodiment, a maintenance and update device for an existing business system is presented in the form of a functional unit. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, or other devices that can provide the above functions.

[0088] Please see Figure 6 , Figure 6 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application, such as... Figure 6 As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 6 Take a processor 10 as an example.

[0089] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0090] The memory 20 stores instructions executable by at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.

[0091] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0092] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0093] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0094] This application also provides a computer-readable storage medium. The methods described in this application can be implemented in hardware or firmware, or implemented as recordable on a storage medium, or implemented as computer code downloaded over a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and subsequently stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the methods shown in the above embodiments are implemented.

[0095] The apparatus or unit described in the above embodiments can be implemented by a computer chip or entity, or by a product having a certain function. A typical implementation device is a computer. Specifically, the computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0096] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0097] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer devices. 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.

[0098] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus, and computer devices 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.

[0099] 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 1 The function specified in one or more boxes.

[0100] 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.

[0101] 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 a process, method, article, or apparatus. Without further limitation, 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 said element.

[0102] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the apparatus embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0103] The above description is merely an embodiment of this application and is 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.

[0104] Although embodiments of this application have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of this application, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for maintaining and updating an existing business system, characterized in that, The method includes: Acquire multi-source heterogeneous information from existing business systems, wherein the multi-source heterogeneous information includes multiple heterogeneous original information from different technology stacks; Generate a standard semantic fragment of any of the heterogeneous original information, extract business semantics from the standard semantic fields to obtain target business fields, and construct an existing business map based on the target business fields; In response to a business update request initiated based on the existing business map, an update prompt message is generated according to the business update request and the existing business map; Based on the update prompt information, the existing business system is updated, and the existing business map is incrementally updated based on the business update results.

2. The method according to claim 1, characterized in that, The standard semantic fragment for generating any of the heterogeneous original information includes: Based on the information source of the heterogeneous original information, the heterogeneous original information is formatted to obtain initial semantic information; The initial semantic information is denoised, and the denoised initial semantic information is segmented to obtain standard semantic segments of the heterogeneous original information.

3. The method according to claim 1, characterized in that, The standard semantic fragment is subjected to business semantic extraction to obtain the target business fields, including: Generate target prompt text based on the standard semantic fragments and preset prompt templates; The target prompt text is input into the semantic extraction model, and the target business field of the standard semantic fragment is output through the semantic extraction model.

4. The method according to claim 1, characterized in that, The target business fields include business module fields, business process fields, function point fields, and data entity fields; constructing an existing business map based on the target business fields includes: For any target business field of the heterogeneous original information, create a corresponding hierarchical node for each field in the target business field; Multiple types of directed edges are established based on the node dependencies between hierarchical nodes. These multiple types of directed edges include call edges, read / write edges, trigger edges, and association edges. Based on the hierarchical nodes and the various types of directed edges, the existing business map is established.

5. The method according to claim 1 or 4, characterized in that, After constructing the existing business map based on the target business fields, the method further includes: The existing business map is rendered on the interactive interface of the existing business system in a specified form, wherein the specified form is a tree diagram, flowchart or topology diagram. The existing business map is collaboratively inspected on the interactive interface, the existing business map is updated based on the collaborative inspection results, and the collaborative inspection results are stored.

6. The method according to claim 1, characterized in that, Based on the aforementioned business update requirements and the existing business map, the generated update notification information includes: In the existing business map, target nodes related to the business update requirements are identified, and the associated technical information of the target nodes is determined. Based on the associated technical information and the business update requirements, at least one clarification question is generated using a large language model. In response to feedback regarding the clarification issue, the update prompt information is generated based on the feedback information and the associated technical information.

7. The method according to claim 1, characterized in that, Updating the existing business system according to the update prompt information includes: The update prompt message is sent to the target programming model or the target development end to instruct the target programming model or the target development end to update the code library of the existing business system according to the update prompt message, so as to complete the business update of the existing business system.

8. The method according to claim 1, characterized in that, Based on the business update results, incremental updates to the existing business map include: Detect change difference information in the codebase of the existing business system and determine the change type of the change difference information; Based on the change type, determine the corresponding target update operation, and in the existing business map, determine the impact nodes and / or impact directed edges associated with the change difference information; The target update operation is performed on the affected nodes and / or the affected directed edges to complete the incremental update of the existing business map.

9. The method according to claim 1, characterized in that, The method further includes: In response to a service deployment request initiated based on the existing service map, an initial set of nodes related to the service deployment request is determined; Traverse the various types of directed edges connecting any initial node in the initial node set to determine the target node set that meets the service deployment requirements; Extract source data from any target node in the target node set, and generate a service deployment package based on the source data. The service deployment package is used for independent deployment and operation in other system environments.

10. A maintenance and update device for an existing business system, characterized in that, The device includes: An information acquisition unit is used to acquire multi-source heterogeneous information from existing business systems, wherein the multi-source heterogeneous information includes multiple heterogeneous original information from different technology stacks. The map construction unit is used to generate a standard semantic fragment of any of the heterogeneous original information, extract business semantics from the standard semantic fragment to obtain target business fields, and construct an existing business map based on the target business fields. The update confirmation unit is used to respond to a business update request initiated based on the existing business map, and generate update prompt information according to the business update request and the existing business map; The update execution unit is used to perform business updates on the existing business system according to the update prompt information, and to perform incremental updates on the existing business map based on the business update results.

11. A computer device, characterized in that, include: A memory and a processor are interconnected, the memory stores computer instructions, and the processor executes the computer instructions to perform the maintenance and update method for the existing business system as described in any one of claims 1 to 9.

Images