Dynamic twinborn classification method and device for rule management, equipment and storage medium

Abstract

The invention provides a dynamic twinborn classification method and device for rule management, equipment and a storage medium, and the method comprises the steps: obtaining a static storage business rule and a logic classification tree with a hierarchical structure, and creating a copy of the logic classification tree for each user as a corresponding initial actual classification tree; acquiring historical data operated by each user in the logic classification tree, and adjusting the initial actual classification tree by using a machine learning algorithm based on the historical data to obtain a first classification tree; based on user test operation, using a reinforcement learning algorithm to adjust the first classification tree in real time to obtain a second classification tree; the logic classification tree and the second classification tree form a dynamic twinborn model; and performing rule management by using the dynamic twinborn model, controlling the display permission of the second classification tree at the corresponding user side through permission management, and updating the dynamic twinborn model by adopting a synchronization mode. According to the method, the dynamic twin double-model structure for rule management is set, so that the query efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of rule management technology, and in particular to a dynamic twin classification method, device, equipment and storage medium for rule management. Background Art

[0002] Rule classification management is a systematic method for organizing, storing, and retrieving rules. It aims to improve the maintainability of the rule base. Rules are usually divided into different categories according to certain logic or characteristics, and a hierarchical structure is established to enable users to locate the content of the required rules.

[0003] In the related technologies, the current rule classification management often relies on a static hierarchical structure to organize and manage rules, that is, all users retrieve the required rule content based on the officially defined hierarchical structure. However, with the continuous changes in business needs and the continuous expansion of the rule base, the existing static management method has the following defects: changes in business needs and rule content cause the existing classification structure to change frequently, resulting in high maintenance costs and low classification efficiency, and an inability to respond quickly to needs; the static classification structure is manually maintained and adjusted by rule designers and is unique, and the rule base classification structure will not be automatically optimized according to actual user query behavior, resulting in increased usage costs when users frequently retrieve a certain rule, affecting the user experience; different roles in the same organization have different access requirements and permissions to the rule base, but the current display method seen by each user is fixed, which is too single and inflexible; in addition, the traditional rule classification system lacks the ability to change the rule life cycle.

[0004] Based on the above analysis of the development status of this technology field, the existing technology lacks a dynamic twin dual-model structure for rule management, and a solution that combines historical data streams with user behavior feedback mechanisms for dynamic updates. Summary of the Invention

[0005] The purpose of the present invention is to provide a rule-managed dynamic twin classification method, device, equipment and storage medium, aiming to solve the above-mentioned problems in the prior art.

[0006] According to a first aspect of an embodiment of the present invention, a rule-managed dynamic twin classification method is provided, comprising:

[0007] Obtaining a logical classification tree with a hierarchical structure that statically stores business rules, and creating a copy of the logical classification tree for each user as the corresponding initial actual classification tree;

[0008] Obtain historical data of each user's operations in the logical classification tree, and use a machine learning algorithm to adjust the initial actual classification tree based on the historical data to obtain a first classification tree; use a reinforcement learning algorithm to adjust the first classification tree in real time based on user test operations to obtain a second classification tree;

[0009] The logical classification tree and the second classification tree form a dynamic twin model;

[0010] Use the dynamic twin model for rule management, control the display permissions of the second classification tree on the corresponding user end through permission management, and update the dynamic twin model in a synchronous manner.

[0011] According to a second aspect of an embodiment of the present invention, a rule-managed dynamic twin classification device is provided, comprising:

[0012] An initial creation module is used to obtain a logical classification tree with a hierarchical structure that statically stores business rules, and to create a copy of the logical classification tree for each user as the corresponding initial actual classification tree;

[0013] The learning and adjustment module is used to obtain the historical data of each user's operation in the logical classification tree, and use the machine learning algorithm to adjust the initial actual classification tree based on the historical data to obtain a first classification tree; based on the user's test operation, the reinforcement learning algorithm is used to adjust the first classification tree in real time to obtain a second classification tree;

[0014] A model storage module is used to combine the logical classification tree and the second classification tree into a dynamic twin model;

[0015] The application update module is used to use the dynamic twin model for rule management, control the display permissions of the second classification tree on the corresponding user end through permission management, and update the dynamic twin model in a synchronous manner.

[0016] According to a third aspect of an embodiment of the present invention, an electronic device is provided, comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the computer program is executed by the processor, the steps of the rule-managed dynamic twin classification method as provided in the first aspect of the present disclosure are implemented.

[0017] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, on which an implementation program for information transmission is stored. When the program is executed by a processor, the steps of the rule-managed dynamic twin classification method provided in the first aspect of the present disclosure are implemented.

[0018] The technical solution provided by the embodiment of the present invention includes the following beneficial effects: defining a dynamic twin architecture DTCM consisting of a logical classification tree and an actual classification tree, the static logical classification tree is used for official staff to update the rule base, and the dynamic actual classification tree can be updated according to user behavior, combined with machine learning of historical data and online machine learning adjustment of real-time data, to provide users with personalized rule management solutions and improve query efficiency.

[0019] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate one or more embodiments of this specification or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0021] Figure 1 is a flow chart of a dynamic twin classification method for rule management according to an embodiment of the present invention;

[0022] Figure 2 is a schematic diagram of a classification tree in an embodiment of the present invention;

[0023] Figure 3 is a schematic diagram of a structure creation code according to an embodiment of the present invention;

[0024] Figure 4 is a schematic diagram of intra-layer mobility according to an embodiment of the present invention;

[0025] Figure 5 Schematic diagram of an overall upward movement of nodes according to an embodiment of the present invention;

[0026] Figure 6 Schematic diagram of node splitting classification according to an embodiment of the present invention;

[0027] Figure 7 is a schematic diagram of the overall architecture of an embodiment of the present invention;

[0028] Figure 8 is a schematic diagram of a rule-managed dynamic twin classification device according to an embodiment of the present invention;

[0029] Figure 9 is a schematic diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0030] In order to enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification will be clearly and completely described below in conjunction with the drawings in one or more embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this document.

[0031] Method Example

[0032] According to an embodiment of the present invention, a dynamic twin classification method for rule management is provided. Figure 1 This is a flow chart of the dynamic twin classification method for rule management according to an embodiment of the present invention. Figure 1 As shown, the rule-managed dynamic twin classification method according to an embodiment of the present invention specifically includes:

[0033] In step S110, a logical classification tree with a hierarchical structure that statically stores business rules is obtained, and a copy of the logical classification tree is created for each user as the corresponding initial actual classification tree, specifically including:

[0034] The classification tree is displayed in a hierarchical structure, reflecting the business logic, hierarchical relationships, and inherent content of the rules. The tree structure organizes classification nodes based on the relationship between business requirements and rules, such as rule type, rule purpose, and scope of application. Each node has a unique parent node and multiple child nodes.

[0035] The static logical classification tree reflects the system's initial design architecture or standardized rule classification, and is developed by relevant enterprise personnel. Retaining the static logical classification tree facilitates official enterprise modification, while the dynamic actual classification tree can adjust the classification hierarchy based on historical user feedback and real-time data. Therefore, there is a standard version of the logical classification tree, and a corresponding version of the actual classification tree for each user terminal.

[0036] In the initial creation phase of the dynamic hierarchical structure, a copy of the logical classification tree is used as the initial actual classification tree corresponding to each user.

[0037] In the embodiment of the present invention, the nodes of the logical classification tree include but are not limited to the following: design stage category: preliminary design, detailed design, construction design, commissioning design, operation and maintenance design; design content category: photovoltaic module layout, inverter selection, grounding system design, bracket design, electrical system design; design specification category: national standards, industry standards, local standards;

[0038] Figure 2 is a schematic diagram of a classification tree in an embodiment of the present invention, such as Figure 2 As shown, some nodes of the logical classification tree are shown as examples. It should be noted that this is only a partial excerpt of the whole. Figure 3 Schematic diagram of the structure creation code of an embodiment of the present invention, such as Figure 3 As shown, the procedure for creating a large class is demonstrated.

[0039] In step S120, historical data of each user's operation in the logical classification tree is obtained, and an initial actual classification tree is adjusted once using a machine learning algorithm based on the historical data to obtain a first classification tree; and a reinforcement learning algorithm is used to adjust the first classification tree in real time based on the user's test operation to obtain a second classification tree, specifically including:

[0040] For each user, it is necessary to learn their historical operation patterns and obtain the name, access frequency, and average duration of each node in the user's access logic classification tree, as well as the complete path length from the root node to the access node as historical data through the user's corresponding unique identifier;

[0041] Convert historical data into numerical feature vectors, and divide the numerical feature vectors into sets based on the length of the complete path. The same complete path length means that the nodes are at the same level in the tree structure. This greatly reduces the possibility of errors in initial adjustments between the same level.

[0042] Use the k-means clustering algorithm to divide each set into clusters;

[0043] Starting from the bottom layer of the initial actual classification tree, we adjust it upward layer by layer. Based on the cluster division results, we use a preset number of moves for each layer. During each move, we randomly obtain node pairs in the same cluster and move one node in the node pair to the parent node of the other node, that is, move at the same layer. We try to make nodes with the same cluster and characteristics belong to the same parent node. Figure 4 Schematic diagram of same-layer mobility according to an embodiment of the present invention. Figure 4 The figure shows the effect of moving on the same layer, and letters are used to distinguish the content of the rules.

[0044] User test operation data is acquired in real time through data stream processing technology. In the embodiment of the present invention, the data stream processing technology includes Apache Kafka or Flink. User test operation refers to a simulated operation before the official provision of the service to avoid unstable adjustments during the official application.

[0045] Based on user test operation data, the optimization of the tree structure can be defined as a multi-step modification, using Q learning or deep Q network as a reinforcement learning algorithm. The state defined by the reinforcement learning algorithm is the tree structure information, and the actions include node movement within the same layer, overall node upward movement, and node splitting and classification. The rewards include the average search time and the length of the complete path. The present invention does not limit other details of the reinforcement learning algorithm, such as the initial value of the Q value in Q learning, and conventional settings are all within the scope of protection of the present invention.

[0046] Same-layer movement refers to same-layer movement in machine learning algorithms. Moving a node as a whole upwards means moving it and all its child nodes to the previous layer as a new classification, setting the previous parent node as its sibling node to avoid hierarchy loss. Node splitting classification means creating a new classification for some of the child nodes under a node, while keeping the rest of the structure unchanged.

[0047] Figure 5 Schematic diagram of the overall upward movement of nodes in an embodiment of the present invention. Figure 5As shown, it shows the effect of overall upward movement. Figure 6 Schematic diagram of node splitting classification according to an embodiment of the present invention. Figure 6 As shown, the effect of split classification is demonstrated.

[0048] The reinforcement learning algorithm can automatically and gradually optimize the tree structure based on the accumulated data of user behavior. For example, if a team is primarily responsible for the design of the photovoltaic panel layout, the reinforcement learning effect is that the photovoltaic panel content may be separated or moved up in the team interface, improving search efficiency and reducing user query time.

[0049] Preferably, since the action selection of adjusting the tree structure is very large, in practice not all actions are worth considering. Obviously invalid actions can be excluded based on domain knowledge or heuristic rules. For example, if the access frequency of a node is determined to be extremely low based on the threshold differentiation method, it will not be considered to be moved to a higher level.

[0050] Preferably, before reinforcement learning, technical means such as calculating the average branching factor, balance factor, entropy value and other indicators can be used to judge the complexity of the first classification tree. If the complexity exceeds a certain threshold, the actions in the reinforcement learning can be controlled to move only on the same layer. That is, when the tree structure is complex, its structure should not be modified randomly.

[0051] The machine learning algorithm based on historical data can quickly form a relatively reasonable tree structure, improve initial performance, and effectively adjust and alleviate the system's cold start problem. Compared with directly starting online learning, it can reduce the computational burden and focus more on data updates and changes during the reinforcement learning process.

[0052] In step S130, the logical classification tree and the second classification tree are combined into a dynamic twin model, specifically including: the dynamic twin model can be defined as DTCM, and the dynamic twin model is stored in a relational database or a graph database, a relational database such as MySQL or PostgreSQL, and a graph database such as Neo4j.

[0053] In step S140, the dynamic twin model is used for rule management, the display rights of the second classification tree on the corresponding user terminal are controlled through rights management, and the dynamic twin model is updated in a synchronous manner, specifically including:

[0054] Identify the user role through the unique identifier corresponding to the user, and determine the display permission of the second classification tree for the user role based on the permission management rules. In the embodiment of the present invention, the user roles include administrators, operators, developers, etc.

[0055] Hide the nodes without display permission and store them in an inaccessible list. Use the second classification tree after permission management control as the background for user-side query. The second classification tree after permission management control can be used as the order of page display or as a query path when there is no page.

[0056] Preferably, the user can customize the display order on the page by adjusting the display weight of the class, but the actual storage of the second classification tree does not change, so as to achieve personalized customized display.

[0057] Use event-driven synchronization and timed synchronization mechanisms as synchronization methods;

[0058] When the logical classification tree in the dynamic twin model changes, the second classification tree is updated synchronously using event-driven synchronization. This is achieved through database trigger monitoring event-driven methods, such as message queue Kafka or synchronization tools such as ETL. Maintenance of inaccessible lists can assist in synchronization detection, otherwise the existence of hidden content may not be determined.

[0059] A timed synchronization mechanism is used to check consistency at preset time intervals to avoid problems with event-driven synchronization.

[0060] The above technical solutions of the embodiments of the present invention are illustrated with reference to the following drawings.

[0061] Figure 7 is a schematic diagram of the overall architecture of an embodiment of the present invention, such as Figure 7 As shown in the figure, it shows the complete process of providing personalized services to users, including preliminary adjustments of machine learning predictions, further optimization by reinforcement learning, and updating and adjusting the classification structure.

[0062] To sum up, in response to the existing problems, the dynamic twin classification method for rule management invented in this invention defines a dynamic twin architecture DTCM consisting of a logical classification tree and an actual classification tree. The static logical classification tree is used for official staff to update the rule base, and the dynamic actual classification tree can be updated according to user behavior; combined with machine learning of historical data and online machine learning adjustment of real-time data, it provides users with personalized rule management solutions and improves query efficiency; the action of modifying the tree during the online learning of real-time data is related to the scale status of the hierarchical structure, so that the dynamic update process conforms to the actual situation; different user groups are distinguished through authority management control, and the distinction is only made after the final form of the second classification tree, that is, the actual tree, is obtained, reducing the logical judgment redundancy of the entire solution; the synchronization method is used to ensure that when the logical classification tree is updated, the dynamic actual classification tree structure is updated accordingly; the overall solution can be widely used in industries with rapid changes and multi-role requirements, such as design, laws and regulations, finance, medical care, and enterprise informatization.

[0063] Device embodiment

[0064] According to an embodiment of the present invention, a rule-managed dynamic twin classification device is provided. Figure 8 Schematic diagram of a dynamic twin classification device for rule management according to an embodiment of the present invention. Figure 8 As shown, the rule-managed dynamic twin classification device according to an embodiment of the present invention specifically includes:

[0065] An initial creation module 80 is used to obtain a logical classification tree with a hierarchical structure that statically stores business rules, and to create a copy of the logical classification tree for each user as a corresponding initial actual classification tree;

[0066] The learning and adjustment module 82 is used to obtain historical data of each user's operation in the logical classification tree, and adjust the initial actual classification tree once based on the historical data using a machine learning algorithm to obtain a first classification tree; and adjust the first classification tree in real time based on user test operations using a reinforcement learning algorithm to obtain a second classification tree. Specifically, it is used to:

[0067] The name, access frequency and average duration of each node in the user's access logic classification tree, and the length of the complete path from the root node to the access node are obtained as historical data through the unique identifier corresponding to the user.

[0068] Convert historical data into numerical feature vectors, divide the numerical feature vectors into sets according to the complete path length, and use the k-means clustering algorithm to divide each set into clusters;

[0069] Starting from the bottom layer of the initial actual classification tree, the tree is adjusted upward layer by layer. Based on the cluster division results, a preset number of moves are used for each layer adjustment. Each move randomly obtains node pairs in the same cluster and moves one node in the node pair to the parent node of the other node.

[0070] Acquire user test operation data in real time through data stream processing technology;

[0071] Based on user test operation data, Q learning or deep Q network is used as the reinforcement learning algorithm. The state defined by reinforcement learning is tree structure information. Actions include moving nodes on the same layer, moving nodes upward as a whole, and splitting and classifying nodes. The rewards include the average search time and the length of the complete path.

[0072] The model storage module 84 is used to combine the logical classification tree and the second classification tree into a dynamic twin model, and is specifically used to store the dynamic twin model in a relational database or a graph database.

[0073] The application update module 86 is used to use the dynamic twin model for rule management, control the display rights of the second classification tree on the corresponding user terminal through permission management, and update the dynamic twin model in a synchronous manner, specifically for:

[0074] The user role is identified by the unique identifier corresponding to the user, and the display permission of the second classification tree for the user role is determined based on the permission management rules. The nodes without display permission are hidden and stored as an inaccessible list, and the second classification tree after permission management control is used as the background for user-side queries.

[0075] Use event-driven synchronization and timed synchronization mechanisms as synchronization methods;

[0076] When the logical classification tree in the dynamic twin model changes, the second classification tree is updated using event-driven synchronization, and the consistency is checked at preset time intervals using a timed synchronization mechanism.

[0077] To sum up, in response to the existing problems, the dynamic twin classification device for rule management invented in this invention defines a dynamic twin architecture DTCM composed of a logical classification tree and an actual classification tree. The static logical classification tree is used for official staff to update the rule base, and the dynamic actual classification tree can be updated according to user behavior; combined with machine learning of historical data and online machine learning adjustment of real-time data, it provides users with personalized rule management solutions and improves query efficiency; the action of modifying the tree during the online learning of real-time data is related to the scale status of the hierarchical structure, so that the dynamic update process conforms to the actual situation; different user groups are distinguished through authority management control, and the distinction is made only after the final form of the second classification tree, that is, the actual tree, is obtained, reducing the logical judgment redundancy of the entire solution; the synchronization method is used to ensure that when the logical classification tree is updated, the dynamic actual classification tree structure is updated accordingly; the overall solution can be widely used in industries with rapid changes and multi-role requirements, such as design, laws and regulations, finance, medical care, and enterprise informatization.

[0078] Electronic device embodiment

[0079] Figure 9 is a schematic diagram of an electronic device according to an embodiment of the present invention. Electronic device 900 may include at least one processor 910 and memory 920. Processor 910 can execute instructions stored in memory 920. Processor 910 is communicatively coupled to memory 920 via a data bus. In addition to memory 920, processor 910 may also be communicatively coupled to input device 930, output device 940, and communication device 950 via the data bus.

[0080] The processor 910 may be any conventional processor, such as a commercially available CPU. The processor may also include a graphics processing unit (GPU), a field programmable gate array (FPGA), a system on chip (SOC), an application specific integrated circuit (ASIC), or a combination thereof.

[0081] The memory 920 may be implemented by any type of volatile or non-volatile memory device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0082] In an embodiment of the present disclosure, executable instructions are stored in the memory 920, and the processor 910 can read the executable instructions from the memory 920 and execute the instructions to implement all or part of the steps of the rule-managed dynamic twin classification method of any of the above exemplary embodiments.

[0083] Computer readable storage medium embodiments

[0084] In addition to the above-mentioned methods and devices, the exemplary embodiments of the present disclosure may also be a computer program product or a computer-readable storage medium storing the computer program product, wherein the computer product includes computer program instructions that can be executed by a processor to implement all or part of the steps described in any of the rule-managed dynamic twin classification methods in the above-mentioned exemplary embodiments.

[0085] The computer program product may be written in any combination of one or more programming languages to write program code for performing the operations of the embodiments of the present application, including object-oriented programming languages such as Java, C++, etc., as well as conventional procedural programming languages such as "C" or similar programming languages and scripting languages (e.g., Python). The program code may be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0086] Computer-readable storage media can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or components, or any combination thereof. More specific examples of readable storage media include: static random access memory (SRAM) electrically connected with one or more wires, electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk, or any suitable combination thereof.

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A rule-managed dynamic twin classification method, characterized in that: include: Obtaining a logical classification tree that statically stores business rules and has a hierarchical structure, and creating a copy of the logical classification tree for each user as a corresponding initial actual classification tree; Obtaining historical data of operations of each user in the logical classification tree, and adjusting the initial actual classification tree once using a machine learning algorithm based on the historical data to obtain a first classification tree; Based on user test operations, the first classification tree is adjusted in real time using a reinforcement learning algorithm to obtain a second classification tree; Combining the logical classification tree and the second classification tree into a dynamic twin model; The dynamic twin model is used for rule management, the display authority of the second classification tree on the corresponding user terminal is controlled through authority management, and the dynamic twin model is updated in a synchronous manner.

2. The method according to claim 1, characterized in that The acquiring of the historical data of each user's operation in the logical classification tree specifically includes: The name, access frequency and average duration of each node in the user's access logic classification tree, and the length of the complete path from the root node to the access node are obtained as historical data through the unique identifier corresponding to the user.

3. The method according to claim 1, characterized in that The adjusting the initial actual classification tree once using a machine learning algorithm based on the historical data to obtain a first classification tree specifically includes: Converting the historical data into numerical feature vectors, dividing the numerical feature vectors into sets according to the complete path length, and using a k-means clustering algorithm to divide the sets into clusters; Starting from the bottom layer of the initial actual classification tree, the tree is adjusted upward layer by layer. Based on the cluster division result, a preset number of moves are used for each layer adjustment. During each move, a node pair in the same cluster is randomly obtained, and one node in the node pair is moved to the parent node of the other node.

4. The method according to claim 1, wherein The method of adjusting the first classification tree in real time based on the user test operation using a reinforcement learning algorithm to obtain the second classification tree specifically includes: Acquire user test operation data in real time through data stream processing technology; Based on the user test operation data, Q learning or deep Q network is used as a reinforcement learning algorithm, wherein the state defined by the reinforcement learning algorithm is tree structure information, the actions include node movement on the same layer, overall node upward movement, and node splitting and classification, and the rewards include the average search time and the complete path length.

5. The method according to claim 1, wherein The step of forming the logical classification tree and the second classification tree into a dynamic twin model specifically includes: storing the dynamic twin model in a relational database or a graph database.

6. The method according to claim 1, characterized in that The controlling the display authority of the second classification tree on the corresponding user terminal through authority management specifically includes: The user role is identified by the unique identifier corresponding to the user, and the display permission of the second classification tree for the user role is determined based on the permission management rules. The nodes without display permission are hidden and stored as an inaccessible list, and the second classification tree after permission management control is used as the background for user-side queries.

7. The method according to claim 1, characterized in that The synchronous updating of the dynamic twin model specifically includes: Use event-driven synchronization and timed synchronization mechanisms as synchronization methods; When the logical classification tree in the dynamic twin model changes, the event-driven synchronization is used to update the second classification tree, and the timing synchronization mechanism is used to check consistency at preset time intervals.

8. A rule-managed dynamic twin classification device, characterized in that: include: An initial creation module is used to obtain a logical classification tree with a hierarchical structure that statically stores business rules, and to create a copy of the logical classification tree for each user as a corresponding initial actual classification tree; A learning and adjustment module is configured to obtain historical data of operations of each user in the logical classification tree, and adjust the initial actual classification tree once using a machine learning algorithm based on the historical data to obtain a first classification tree; Based on user test operations, the first classification tree is adjusted in real time using a reinforcement learning algorithm to obtain a second classification tree; A model storage module, configured to combine the logical classification tree and the second classification tree into a dynamic twin model; An application update module is used to use the dynamic twin model for rule management, control the display permissions of the second classification tree on the corresponding user terminal through permission management, and update the dynamic twin model in a synchronous manner.

9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the steps of the rule-managed dynamic twin classification method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores an implementation program for information transmission, and when the program is executed by a processor, the steps of the rule-managed dynamic twin classification method according to any one of claims 1 to 7 are implemented.

Images