A digital management method for enterprise internal control based on AI big model

Abstract

The present invention discloses a digital management method for enterprise internal control based on an AI big model, which relates to the field of enterprise data management technology and is used to solve the problem in the prior art that enterprise data is difficult to be supervised when anomalies occur, and the security of data interaction between enterprises is poor. It is achieved by creating an internal control data warehouse for each department within the enterprise, obtaining the enterprise operation data of each department for digital processing, obtaining the digital data to be controlled by each enterprise, and storing it in the respective internal control data warehouses, constructing an AI big model for supervising the enterprise operation data corresponding to all departments of the enterprise, analyzing the abnormal data within the corresponding enterprise of each department based on the AI big model, and performing internal control. When data interacts between different enterprises, it is judged whether there is direct handover authority between the two, and based on the judgment result, internal related data interaction or external transit data interaction is selected, and the interactive flow data of the two enterprises interacting with data is fully controlled.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise data management, and specifically to a digital management method for enterprise internal control based on an AI big model. Background Art

[0002] AI big models refer to deep learning models with huge parameter levels, usually reaching billions or even trillions. They are trained based on massive data and can achieve multi-task and multi-modal intelligent analysis and generation through complex neural network architecture.

[0003] In existing enterprise management, the operational data corresponding to each department lacks effective supervision, security is difficult to guarantee, and the discovery of problematic data is not timely and accurate enough. The emergence of AI big models provides an effective means for data supervision. How to use AI big models to achieve data management of different departments under each enterprise, and how to supervise data interaction between different enterprises to ensure the security of data interaction, are issues that urgently need to be solved. Summary of the Invention

[0004] In order to solve the above problems, the purpose of the present invention is to provide a digital management method for enterprise internal control based on AI big model.

[0005] The purpose of the present invention can be achieved through the following technical solution: A digital management method for enterprise internal control based on AI big model, comprising the following steps:

[0006] Step S1: Create an internal control data warehouse for each department within the enterprise, obtain the enterprise operation data corresponding to each department, and digitize the enterprise operation data to obtain the digital data to be controlled corresponding to each department, and store it in the respective internal control data warehouse;

[0007] Step S2: Build an AI big model to monitor the enterprise's operational data for all departments. Based on the AI big model, analyze and obtain abnormal data within each department, and conduct internal control on the abnormal data.

[0008] Step S3: When data is exchanged between different enterprises, determine whether there is direct handover authority between the enterprises, and based on the judgment result, choose to conduct internal related data interaction or external transit data interaction, and conduct full-process management and control of the interactive flow data of the two enterprises interacting with the data.

[0009] Furthermore, the process of creating an internal control data warehouse for each department within the enterprise includes:

[0010] All departments within the enterprise are numbered in sequence and denoted as i, i = 1, 2, 3, ..., n, where n is a natural number greater than 0. Each department under the enterprise is associated with a corresponding authorization identification code, which is used to determine the authorization status of the corresponding department;

[0011] Build data access layer, data cleaning rule base and permission unit;

[0012] Authorization status includes "Authorized" and "Unauthorized";

[0013] When a department in an enterprise is in the "authorized" state, it initiates a database construction application instruction with the total data storage of the current department and transmits it to the data access layer. Based on the total data storage of the corresponding department's data, the data access layer creates corresponding memory space to build the preliminary data warehouse of the corresponding enterprise;

[0014] The preliminary data warehouse consists of several data blocks. A monitoring cycle is set for each data block. The environmental information of each data block during the monitoring cycle is cleaned using a data cleaning rule library. The data warehouse is assigned permissions to the permission unit, allocating internal control permissions for data interaction between different departments within the enterprise. These permissions are then associated with the preliminary data warehouse corresponding to the corresponding department, creating an internal control data warehouse for each department.

[0015] When a department under an enterprise is in "unauthorized" status, no operations will be performed.

[0016] Furthermore, the process of obtaining the enterprise operation data corresponding to each department, digitizing the enterprise operation data, obtaining the digital data to be controlled corresponding to each department, and storing it in the respective internal control data warehouse includes:

[0017] Set data entry periods for each department within the enterprise. Each department will obtain its own enterprise operation data. Enterprise operation data includes financial data, transaction data, and inventory data from the finance department; document data, image data, and audio and video data from the sales department; equipment operation data, business indicator data, and server log data from the technical operation and maintenance department; and customer data and supplier data from the operations department.

[0018] Configure digital processing tools, and use them to digitally process the enterprise operation data of each department under the enterprise according to the configured data processing format, data processing rules and verification rules, to obtain the digital data to be controlled corresponding to each department, and store the digital data to be controlled of each department into the corresponding internal control data warehouse, and monitor the storage environment. When an abnormality occurs in the storage environment, stop the storage of the digital data to be controlled corresponding to the corresponding department.

[0019] Furthermore, the process of building an AI-powered model to monitor the operational data of all departments within an enterprise includes:

[0020] Construct a hierarchical model architecture corresponding to the AI large-scale model that oversees the enterprise operation data of various departments. The specific components of the hierarchical model architecture include a multi-source data input layer, a modal encoding layer, a cross-modal alignment layer, a domain knowledge injection layer, a task decision layer, and a feedback optimization layer.

[0021] Model fusion is performed according to the hierarchical model architecture to build an AI big model that ultimately supervises the enterprise operation data of each department under the enterprise. The historical operation data corresponding to each department is obtained, and the AI big model is trained after dividing the training set, verification set and test set until the model prediction rate of the AI big model meets the preset standard threshold.

[0022] Furthermore, the AI model is used to analyze abnormal data within each department of the enterprise and conduct internal control of abnormal data. The process includes:

[0023] The constructed AI model analyzes the digital data to be controlled in each department in turn. If the analysis finds abnormal data in the digital data to be controlled in a certain department, the storage path and abnormal items of the abnormal data are further located.

[0024] Construct an internal control unit and perform internal control on the abnormal data located by each department in turn. Internal control includes traversing the data from the root node of the abnormal data in the storage path, marking the abnormal positions of all abnormal items traversed on the storage path, creating an internal control sub-node at each abnormal position on the storage path, editing the abnormal repair script, and injecting it into each internal control sub-node on the storage path in the form of code. The internal control sub-node executes the modification of the corresponding abnormal item at each abnormal position according to the abnormal repair script. If the modification of the abnormal item cannot be completed, the corresponding abnormal item will be eliminated.

[0025] Furthermore, when different enterprises interact with each other, the process of determining whether there is direct handover authority between the enterprises and selecting whether to interact with internal related data or external transit data based on the determination result includes:

[0026] When two companies interact with each other, they each create an interaction request.

[0027] Build an interaction rule matching library, which stores several enterprise pairing relationship groups. An enterprise pairing relationship group is used to include two enterprises with business interaction needs, as well as data interaction between the respective departments of the two enterprises;

[0028] There is direct handover authority between the two enterprises in the enterprise pairing relationship group, and the interaction requests of the two enterprises in the enterprise pairing relationship group are associated in the interaction rule matching library;

[0029] When it is determined that two enterprises that are interacting with data have direct handover authority, internal related data interaction is performed for the data interaction between the two enterprises. When the two enterprises do not have direct handover authority, external transit data interaction is performed for the data interaction between the two enterprises.

[0030] Furthermore, internal related data interaction and external transit data interaction each include:

[0031] Internal data interaction involves building internal data transfer channels between enterprises. When departments of the same type in different enterprises interact with each other, the data can be directly transferred through the internal data transfer channels.

[0032] External transit data interaction is as follows: an intermediate cache component is built within the internal data flow channel. When different types of departments under two enterprises interact with each other, each department transfers the relevant data that needs to be interacted with to the intermediate cache component. The intermediate cache component acts as a data transfer party to interact with the relevant data of the two enterprises and saves all the data of the data interaction in the intermediate cache component.

[0033] Furthermore, the process of full-process control of the interactive data flow between two companies includes:

[0034] When two enterprises interact with each other, the interactive flow data generated during the data interaction is mapped to the pre-built control chain, and a flow timeline is constructed for the flow process of the interactive flow data between the two enterprises. The flow timeline includes several flow time nodes.

[0035] Mark the timestamp of the interactive flow data at each flow time node, and analyze whether there is any abnormal data flow operation at each timestamp. If so, suspend the data interaction between the two enterprises, analyze the cause of the abnormal data flow operation, and obtain the corresponding troubleshooting measures for the abnormal data flow operation. If not, do not perform any operation.

[0036] Compared with the existing technology, the beneficial effects of the present invention are: by creating a corresponding internal control data warehouse for each department within the enterprise, obtaining the enterprise operation data of each department for digital processing, obtaining the respective digital data to be controlled and storing them in their respective internal control data warehouses, building an AI large model for supervising enterprise operation data, analyzing and obtaining abnormal data within each department corresponding to the enterprise, and conducting internal control. When data is interacted between different enterprises, it is judged whether there is direct handover authority between the enterprises, and based on the judgment result, it is selected to interact with internal related data or interact with external transit data, and the interactive flow data is controlled throughout the process, which solves the current problem that enterprise data is difficult to supervise when abnormalities occur and the security of data interaction between enterprises is poor. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 Flowchart of the present invention. DETAILED DESCRIPTION

[0038] like Figure 1 As shown, a digital management method for enterprise internal control based on an AI big model includes the following steps:

[0039] Step S1: Create an internal control data warehouse for each department within the enterprise, obtain the enterprise operation data corresponding to each department, and digitize the enterprise operation data to obtain the digital data to be controlled corresponding to each department, and store it in the respective internal control data warehouse;

[0040] Step S2: Build an AI big model to monitor the enterprise's operational data for all departments. Based on the AI big model, analyze and obtain abnormal data within each department, and conduct internal control on the abnormal data.

[0041] Step S3: When data is exchanged between different enterprises, determine whether there is direct handover authority between the enterprises, and based on the judgment result, choose to conduct internal related data interaction or external transit data interaction, and conduct full-process management and control of the interactive flow data of the two enterprises interacting with the data.

[0042] It should be further explained that, in the specific implementation process, the process of creating an internal control data warehouse for each department within the enterprise includes:

[0043] All departments in the enterprise are numbered in sequence, and the number is recorded as i, then i = 1, 2, 3, ..., n, where n is a natural number greater than 0. Each department in the enterprise is associated with a corresponding authorization identification code, which is used to determine the authorization status of the corresponding department;

[0044] Build data access layer, data cleaning rule base and permission unit;

[0045] Starting from label i=1, the authorization identification code of each department is traversed in turn. According to the authorization identification code, whether the corresponding department belongs to the preset permission table is determined. If the authorization identification code is in the permission table, the authorization status of the corresponding department is determined to be "authorized". If the authorization identification code is not in the permission table, the authorization status of the corresponding department is determined to be "unauthorized";

[0046] When a department in an enterprise is in the "authorized" state, it initiates a database construction application instruction with the total data storage of the current department and transmits the instruction to the data access layer. After receiving the instruction, the data access layer creates the corresponding memory space in the data access layer according to the total data storage of the corresponding data of the corresponding department, and then builds the preliminary data warehouse of the corresponding enterprise;

[0047] The preliminary data warehouse consists of several data blocks. A monitoring period is set for each data block. The environmental information of each data block during the monitoring period is cleaned using a data cleaning rule library. The alienated interference information corresponding to each data block is screened out and removed. The presence of such alienated interference information will have a negative impact on the storage environment of the corresponding data block and interfere with the normal operation of data storage.

[0048] After removing all the alienated and interfering information from all data blocks, continue to perform the permission allocation operation on the permission unit of the data warehouse, allocate internal control permissions for data interaction between different departments of the enterprise, and associate the internal control permissions with the corresponding preliminary data warehouse of the corresponding department, thereby completing the creation of the internal control data warehouse corresponding to each department;

[0049] The internal control authority is explained as follows: when different departments of an enterprise interact with each other, the data interaction channel is within the internal interaction channel pre-set by the enterprise, that is, data interaction between departments is carried out through the internal interaction channel;

[0050] When a department under an enterprise is in the "unauthorized" authorization state, no operation will be performed.

[0051] It should be further explained that, during the specific implementation process, the process of obtaining the enterprise operation data corresponding to each department, digitizing the enterprise operation data, obtaining the corresponding digital data to be controlled by each department, and storing it in the respective internal control data warehouse includes:

[0052] Set data entry periods for each department within the enterprise. During these periods, each department will obtain its own enterprise operation data. This data includes financial data, transaction data, and inventory data from the finance department; document data, image data, and audio and video data from the sales department; equipment operation data, business indicator data, and server log data from the technical operations department; and customer data and supplier data from the operations department.

[0053] The data of the financial department are described as follows:

[0054] The financial data includes but is not limited to accounting documents (debit / credit), general ledger details, and balance sheets; the transaction data includes but is not limited to sales orders (SKU / quantity / amount), purchase contracts (terms / delivery date), and transaction contracts (transaction date / transaction party); the inventory data includes but is not limited to real-time inventory levels (warehouse / shelf level), turnover rates, and safety stock thresholds;

[0055] The data of the sales department are described as follows:

[0056] Document data includes but is not limited to PDF contracts, Word technical agreements, and Excel quotations; image data includes but is not limited to product design drawings, quality inspection photos, and sales receipts; and audio and video data includes but is not limited to customer service call recordings and production site surveillance videos;

[0057] The data of the technical operation and maintenance department are described as follows:

[0058] The equipment operation data includes but is not limited to equipment operating temperature, equipment vibration frequency and equipment working energy consumption; the business indicator data includes but is not limited to website real-time UV, website real-time PV and equipment comprehensive efficiency; the server log data includes but is not limited to server access concurrent data volume and server response time;

[0059] The data of the operating department are described as follows:

[0060] The customer data includes, but is not limited to, basic information such as customer classification (A, B, and C), credit limit, and historical transaction amount, as well as behavioral data such as website click-through rate and customer service consultation subject. The supplier data includes, but is not limited to, the quality qualification rate of delivered products, on-time delivery rate, and cost competitiveness rate;

[0061] Configure digital processing tools to digitally process the enterprise operation data corresponding to each department of the enterprise according to the configured data processing format, data processing rules, and verification rules, and then obtain the digital data to be controlled corresponding to each department;

[0062] The digital data to be controlled of each department will be stored in the corresponding internal control data warehouse, and the storage environment will be monitored. When an abnormality occurs in the storage environment, the storage of the corresponding digital data to be controlled by the corresponding department will be stopped.

[0063] It should be further explained that, in the specific implementation process, the process of building an AI model to supervise the enterprise operation data of all departments includes:

[0064] Construct a hierarchical model architecture corresponding to the AI large-scale model that oversees the enterprise operation data of various departments. The specific components of the hierarchical model architecture include a multi-source data input layer, a modal encoding layer, a cross-modal alignment layer, a domain knowledge injection layer, a task decision layer, and a feedback optimization layer.

[0065] The multi-source data input layer is used to input the corresponding digital data to be controlled in the internal control data warehouse of each department;

[0066] The modality encoding layer consists of a text encoder, a temporal encoder, and a graph structure encoder;

[0067] The text encoder models the long text of each department's digital data to be managed;

[0068] The time series encoder analyzes the time series dependencies of the digital data to be controlled in each department;

[0069] The graph structure encoder analyzes the graph structure of the digital data to be controlled in each department;

[0070] The cross-modal alignment layer is used to set a cross-modal interaction interface for the modal coding layers corresponding to different departments, and then align the modal coding layers corresponding to each two departments performing data interaction to a unified dimensional level to complete the data interaction between the corresponding departments;

[0071] The domain knowledge injection layer is used to inject the domain knowledge graph corresponding to each department. The domain knowledge graph is used as the supervision indicator corresponding to each department, and the substandard data in each department is located through the supervision indicator;

[0072] The task decision layer is used to make decisions and analyze the data anomaly level corresponding to each department;

[0073] The feedback optimization layer constructs the activation function and loss function, and performs data training on all architecture layers in the hierarchical model architecture except itself. Data training is stopped after each architecture layer reaches its own preset architecture indicator to complete the optimization of each architecture layer.

[0074] Model fusion is performed according to the hierarchical model architecture, and then an AI big model is constructed to ultimately supervise the enterprise operation data of each department under the enterprise. The historical operation data corresponding to each department is obtained, and the AI big model is trained after dividing the training set, verification set and test set until the model prediction rate of the AI big model meets the preset standard threshold.

[0075] It should be further explained that, during the specific implementation process, the process of analyzing abnormal data within each department of the enterprise based on the AI big model and conducting internal control of abnormal data includes:

[0076] The constructed AI model analyzes the digital data to be controlled in each department in turn. If the analysis finds abnormal data in the digital data to be controlled in a certain department, the storage path and abnormal items of the abnormal data are further located.

[0077] Construct an internal control unit, which will perform internal control on the abnormal data located by each department in turn. The internal control includes traversing the data from the root node of the abnormal data in the storage path, marking the abnormal positions of all abnormal items traversed on the storage path, creating an internal control sub-node at each abnormal position on the storage path, editing the abnormal repair script, and injecting the abnormal repair script into the code of each internal control sub-node on the storage path. The internal control sub-node executes the modification of the corresponding abnormal item at each abnormal position according to the abnormal repair script. If the modification of the abnormal item cannot be completed, the corresponding abnormal item will be eliminated.

[0078] It should be further explained that, in the specific implementation process, when different enterprises interact with each other, the process of determining whether there is direct handover authority between the enterprises and selecting whether to interact with internal related data or external transit data based on the determination result includes:

[0079] When two companies interact with each other, they each create an interaction request.

[0080] Build an interaction rule matching library, which stores several enterprise pairing relationship groups. An enterprise pairing relationship group is used to include two enterprises with business interaction needs, as well as data interaction between the respective departments of the two enterprises;

[0081] There is direct handover authority between the two enterprises in the enterprise pairing relationship group, and the interaction requests corresponding to the two enterprises in the enterprise pairing relationship group are associated in the interaction rule matching library;

[0082] The enterprise pairing relationship group is used to pair the relationships between the same departments of different enterprises;

[0083] When it is determined that the two enterprises performing data interaction have direct handover authority, the data interaction between the two enterprises is performed as internal related data interaction; when it is determined that the two enterprises performing data interaction do not have direct handover authority, the data interaction between the two enterprises is performed as external transfer data interaction;

[0084] The content of the internal related data interaction is: building an internal data flow channel between enterprises. When departments of the same type under different enterprises interact with each other, the departments interact directly through the internal data flow channel.

[0085] The content of the external transit data interaction is: building an intermediate cache component in the internal data flow channel. When different types of departments under two enterprises interact with each other, each department will transfer the relevant data that needs to be interacted with to the intermediate cache component. The intermediate cache component acts as a data transfer party to interact with the relevant data of the two enterprises, and saves all the data of the data interaction at the intermediate cache component.

[0086] It should be further explained that, during the specific implementation process, the process of full-process control of the interactive data flow between two companies includes:

[0087] When two companies interact with each other, the interactive flow data generated during the data interaction is mapped to the pre-built control chain, and a corresponding flow timeline is constructed in the control chain for the flow process of the interactive flow data between the two companies.

[0088] The flow time axis includes a plurality of flow time nodes;

[0089] Mark the timestamp of the interactive flow data at each flow time node and analyze whether there are any abnormal data flow operations at each timestamp. Abnormal data flow operations include abnormal data throughput, abnormal data interruption, and abnormal data transmission delay.

[0090] If so, data interaction between the two enterprises is suspended, and the cause of the abnormal data flow operation is analyzed. A corresponding data log form is constructed and uploaded to a pre-built troubleshooting database. The troubleshooting database then obtains the troubleshooting measures corresponding to the abnormal data flow operation between the enterprises currently interacting with data, until the entire process of control over each timestamp of the interactive flow data that is subject to abnormal data flow operation is completed.

[0091] If not, no action is taken.

[0092] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A digital management method for enterprise internal control based on AI big model, characterized by: The following steps are involved: Step S1: Create an internal control data warehouse for each department within the enterprise, obtain the enterprise operation data corresponding to each department, and digitize the enterprise operation data to obtain the digital data to be controlled corresponding to each department, and store it in the respective internal control data warehouse; Step S2: Build an AI big model to monitor the enterprise's operational data for all departments. Based on the AI big model, analyze and obtain abnormal data within each department, and conduct internal control on the abnormal data. Step S3: When data is exchanged between different enterprises, determine whether there is direct handover authority between the enterprises. Based on the determination result, choose to exchange internal related data or external transit data, and perform full-process control over the interactive data flow between the two enterprises. The process of building an AI-powered model to monitor operational data across all departments of an enterprise includes: Construct a hierarchical model architecture corresponding to the AI large-scale model that oversees the enterprise operation data of various departments. The specific components of the hierarchical model architecture include a multi-source data input layer, a modal encoding layer, a cross-modal alignment layer, a domain knowledge injection layer, a task decision layer, and a feedback optimization layer. Based on the hierarchical model architecture, model fusion is performed to build an AI model that ultimately monitors the enterprise operation data of each department within the enterprise. The historical operation data corresponding to each department is obtained, and the AI model is trained after dividing it into training, validation, and test sets until the model prediction rate of the AI model meets the preset threshold. The AI big model analyzes abnormal data within each department of the enterprise and conducts internal control of abnormal data. The process includes: The constructed AI model analyzes the digital data to be controlled in each department in turn. If the analysis finds abnormal data in the digital data to be controlled in a certain department, the storage path and abnormal items of the abnormal data are further located. Construct an internal control unit and perform internal control on the abnormal data located by each department in turn. Internal control includes traversing the data from the root node of the abnormal data in the storage path, marking the abnormal positions of all abnormal items traversed on the storage path, creating an internal control sub-node at each abnormal position on the storage path, editing the abnormal repair script, and injecting it into each internal control sub-node on the storage path in the form of code. The internal control sub-node executes the modification of the corresponding abnormal item at each abnormal position according to the abnormal repair script. If the modification of the abnormal item cannot be completed, the corresponding abnormal item will be eliminated.

2. The enterprise internal control digital management method based on AI big model according to claim 1 is characterized by: The process of creating an internal control data warehouse for each department within the enterprise includes: All departments within the enterprise are numbered in sequence and denoted as i, i = 1, 2, 3, ..., n, where n is a natural number greater than 0. Each department under the enterprise is associated with a corresponding authorization identification code, which is used to determine the authorization status of the corresponding department; Build data access layer, data cleaning rule base and permission unit; Authorization status includes "Authorized" and "Unauthorized"; When a department within an enterprise is in the "Authorized" state, it initiates a database construction request with the department's total data storage capacity and transmits it to the data access layer. Based on the department's total data storage capacity, the data access layer creates the corresponding memory space and builds the preliminary data warehouse for the enterprise. The preliminary data warehouse consists of several data blocks. A monitoring cycle is set for each data block. The environmental information of each data block during the monitoring cycle is cleaned using a data cleaning rule library. The data warehouse is assigned permissions to the permission unit, allocating internal control permissions for data interaction between different departments within the enterprise. These permissions are then associated with the preliminary data warehouse corresponding to the corresponding department, creating an internal control data warehouse for each department. When a department under an enterprise is in "unauthorized" status, no operations will be performed.

3. The enterprise internal control digital management method based on AI big model according to claim 2 is characterized in that: The process of obtaining the enterprise operation data corresponding to each department, digitizing the enterprise operation data, obtaining the corresponding digital data to be controlled by each department, and storing it in the respective internal control data warehouse includes: Set data entry periods for each department within the enterprise. Each department will obtain its own enterprise operation data. Enterprise operation data includes financial data, transaction data, and inventory data from the finance department; document data, image data, and audio and video data from the sales department; equipment operation data, business indicator data, and server log data from the technical operation and maintenance department; and customer data and supplier data from the operations department. Configure digital processing tools, and use them to digitally process the enterprise operation data of each department under the enterprise according to the configured data processing format, data processing rules and verification rules, to obtain the digital data to be controlled corresponding to each department, and store the digital data to be controlled of each department into the corresponding internal control data warehouse, and monitor the storage environment. When an abnormality occurs in the storage environment, stop the storage of the digital data to be controlled corresponding to the corresponding department.

4. The enterprise internal control digital management method based on AI big model according to claim 3 is characterized by: When different enterprises interact with each other, the process of determining whether there is direct handover authority between the enterprises and choosing whether to interact with internal related data or external transit data based on the determination result includes: When two companies interact with each other, they each create an interaction request. Build an interaction rule matching library, which stores several enterprise pairing relationship groups. An enterprise pairing relationship group is used to include two enterprises with business interaction needs, as well as data interaction between the respective departments of the two enterprises; There is direct handover authority between the two enterprises in the enterprise pairing relationship group, and the interaction requests of the two enterprises in the enterprise pairing relationship group are associated in the interaction rule matching library; When it is determined that two enterprises that are interacting with data have direct handover authority, internal related data interaction is performed for the data interaction between the two enterprises. When the two enterprises do not have direct handover authority, external transit data interaction is performed for the data interaction between the two enterprises.

5. The enterprise internal control digital management method based on AI big model according to claim 4 is characterized in that: Internal related data interaction and external transit data interaction each include: Internal data interaction involves building internal data transfer channels between enterprises. When departments of the same type in different enterprises interact with each other, the data can be directly transferred through the internal data transfer channels. External transit data interaction is as follows: an intermediate cache component is built within the internal data flow channel. When different types of departments under two enterprises interact with each other, each department transfers the relevant data that needs to be interacted with to the intermediate cache component. The intermediate cache component acts as a data transfer party to interact with the relevant data of the two enterprises and saves all the data of the data interaction in the intermediate cache component.

6. The enterprise internal control digital management method based on AI big model according to claim 5 is characterized in that: The process of fully managing and controlling the interactive data flow between two companies involves: When two enterprises interact with each other, the interactive flow data generated during the data interaction is mapped to the pre-built control chain, and a flow timeline is constructed for the flow process of the interactive flow data between the two enterprises. The flow timeline includes several flow time nodes. Mark the timestamp of the interactive flow data at each flow time node, and analyze whether there is any abnormal data flow operation at each timestamp. If so, suspend the data interaction between the two enterprises, analyze the cause of the abnormal data flow operation, and obtain the corresponding troubleshooting measures for the abnormal data flow operation. If not, do not perform any operation.

Images