Enterprise operation cost analysis method and device based on privacy calculation, equipment and medium

Abstract

The invention provides an enterprise operation cost analysis method and device based on privacy calculation, equipment and a medium, and belongs to the technical field of data processing.The method comprises the steps that an enterprise data source is determined, and a data collection interface is set to collect, pre-process and encrypt service data; constructing a privacy computing platform, wherein each data source and the privacy computing platform construct a unified enterprise operation cost analysis basic model through a federal learning mode; carrying out encryption calculation among the data sources in a multi-party security calculation mode, and uploading an encryption calculation result to the privacy calculation platform; and the unified enterprise operation cost analysis basic model is optimized by using the encrypted calculation result on the privacy calculation platform, then the optimized unified enterprise operation cost analysis basic model is used for enterprise operation privacy calculation, and different viewing permissions are set for the enterprise operation privacy calculation result. According to the method, the enterprise data privacy is fully protected, and the accuracy and efficiency of operating cost analysis are improved.

Description

Technical Field

[0001] This application belongs to the field of data processing technology and relates to a method, device, equipment and medium for analyzing enterprise operating costs based on privacy computing. Background Art

[0002] As businesses of all sizes expand, operating cost analysis becomes increasingly crucial for business decision-making. Traditional operating cost analysis relies on centralized collection of relevant data from various internal departments and external sources, followed by unified analysis. However, this approach carries serious data privacy risks. For example, data from departments like finance, sales, and production often contains sensitive information such as employee privacy and business secrets. Once aggregated for analysis, data leakage is highly likely.

[0003] Despite the availability of privacy-preserving technologies such as encryption, they continue to encounter numerous problems when applied to enterprise operating cost analysis. Some encryption technologies, after encryption, result in low computational efficiency and lengthy analysis processes, making them incapable of meeting the demands of real-time decision-making. Furthermore, during data interaction and collaborative computing, privacy controls are difficult to implement accurately, creating the constant risk of privacy leaks. While the industry has attempted to address these issues by optimizing encryption algorithms and establishing data access rights management, even optimized encryption algorithms cannot meet the complex computational requirements of enterprise operating cost analysis, and data rights management cannot accommodate cross-departmental and cross-institutional data collaboration scenarios. Therefore, there is currently a lack of an effective solution for enterprise operating cost analysis that balances data privacy protection with the needs for efficient analysis. Summary of the Invention

[0004] In a first aspect, embodiments of the present application provide a method for analyzing enterprise operating costs based on privacy computing, comprising the following steps: S1. Identify the enterprise's internal business systems and external collaborative systems as data sources, and set up data collection interfaces in each data source to collect, pre-process, and encrypt their respective business data; S2. Build a privacy-preserving computing platform. Each data source and the privacy-preserving computing platform use federated learning to construct a unified basic model for enterprise operating cost analysis. S3. Split the data from each data source and send the split parts of the same data source to different data sources, thereby performing encrypted calculations between the data sources through multi-party secure computing, and uploading the encrypted calculation results to the privacy computing platform; S4. Use the encrypted calculation results on the privacy computing platform to optimize the unified basic model for enterprise operating cost analysis, then use the optimized unified basic model for enterprise operating cost analysis to perform enterprise operating privacy calculations, and set different viewing permissions for the enterprise operating privacy calculation results.

[0005] Furthermore, the specific steps of step S1 are as follows: S11. Identify the business systems within the enterprise as internal data sources and the external collaborative systems as external data sources; S12. Set up data collection interfaces for both internal and external data sources; S13. The business data of the corresponding data source is collected through the data collection interface, and is pre-processed according to the respective preset pre-processing methods, and then encrypted according to the preset encryption method.

[0006] Furthermore, the specific steps of step S2 are as follows: S21. Build a privacy-preserving computing platform and establish connections between the privacy-preserving computing platform and various data sources; S22. Filter out the data source required for federated learning from the data source as the first data source, and deploy a federated learning client on the first data source; S23. Each first data source constructs an enterprise operating cost analysis sub-model on its corresponding federated learning client and trains it using its own encrypted business data. After training, the sub-model parameters are obtained and encrypted and uploaded to the privacy computing platform. S24. The privacy computing platform decrypts the parameters of each sub-model and performs weighted calculation according to the set weights to obtain the comprehensive weighted parameters. The comprehensive weighted parameters are returned to the federated learning clients of each first data source for retraining the sub-models until the iteration termination condition is met. S25. The federated learning clients of each first data source return each enterprise operating cost analysis sub-model to the privacy computing platform to obtain a unified enterprise operating cost analysis basic model.

[0007] Furthermore, the specific steps of step S3 are as follows: S31. Filter out the data source required for multi-party secure computing from the data source as the second data source, and determine the number N of the second data source; S32. Split the data in each second data source to ensure that each record is split into N-1 parts; S33. Send the N-1 portion of data from each second data source to a different second data source to ensure that a complete data record of each second data source is not obtained by another second data source; S34. Each second data source calculates an intermediate result for its own partial data and the N-1 partial data obtained from each other second data source according to the logical relationship of the data fields. The intermediate result is encrypted, transmitted, and combined for calculation between the second data sources through the obfuscation circuit to obtain an encrypted calculation result. S35. Each second data source uploads its own encrypted calculation result part to the privacy computing platform.

[0008] Furthermore, the specific steps of step S4 are as follows: S41. The privacy computing platform receives the encrypted calculation results uploaded by each second data source and decrypts them to obtain partial data from each data source; S42. The privacy computing platform uses partial data from various data sources to optimize the parameters of the unified enterprise operating cost analysis basic model until the model meets the requirements. S43. The privacy computing platform obtains privacy computing requirements, parses the encrypted data from the corresponding business system as the data source, and inputs it into the optimized unified enterprise operating cost analysis basic model to complete the enterprise operation privacy computing. S44. The privacy computing platform will determine the privacy level of the enterprise's operational privacy calculation results and, based on the enterprise's internal user role permissions, assign corresponding viewing permissions to different user roles. S45. The privacy computing platform verifies the corresponding permissions of users who view the enterprise operation privacy computing results through multi-factor authentication.

[0009] Furthermore, in step S21, a distributed privacy computing platform is constructed, wherein the privacy computing platform includes a plurality of computing nodes; The privacy computing platform treats both comprehensive weighted parameter calculations and enterprise operation privacy calculations as computing tasks, and distributes the computing tasks to each computing node for parallel execution during the calculation process.

[0010] Furthermore, the method further comprises the following steps: The privacy computing platform records the time when each data source uploads model parameters and receives comprehensive weighted parameters, the sub-model training process of the federated learning client of each data source, the data segmentation in multi-party secure computing, and the calculation of intermediate results into the blockchain.

[0011] In a second aspect, an embodiment of the present application further provides an enterprise operating cost analysis device based on privacy computing, comprising: The data collection module is used to determine the enterprise's internal business systems and external cooperation systems as data sources. A data collection interface is set up in each data source to collect, pre-process and encrypt their respective business data; The privacy computing module is used to build a privacy computing platform. Various data sources and the privacy computing platform use federated learning to build a unified basic model for enterprise operating cost analysis. A multi-party secure computing module is used to split the data of each data source and send the split parts of the same data source to different data sources, thereby performing encrypted calculations between the data sources through multi-party secure computing and uploading the encrypted calculation results to the privacy computing platform; The enterprise operating cost analysis and calculation module is used to optimize the unified enterprise operating cost analysis basic model using encrypted calculation results on the privacy computing platform, and then use the optimized unified enterprise operating cost analysis basic model to perform enterprise operating privacy calculations, and set different viewing permissions for the enterprise operating privacy calculation results.

[0012] In a third aspect, an embodiment of the present application also provides an electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the enterprise operating cost analysis method based on privacy computing as described in the first aspect are implemented.

[0013] In a fourth aspect, an embodiment of the present application further provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the enterprise operating cost analysis method based on privacy computing as described in the first aspect are implemented.

[0014] It can be seen from the above technical solutions that this application has the following advantages: The privacy-preserving enterprise operating cost analysis method, device, equipment, and media provided in this application integrate data collection, federated learning, multi-party secure computing, and privacy-preserving computing. By integrating internal and external data sources, using federated learning to build a unified basic model for enterprise operating cost analysis, optimizing the model with multi-party secure computing, and performing privacy-preserving computing using the optimized model, this method not only fully protects enterprise data privacy but also improves the accuracy and efficiency of operating cost analysis. Furthermore, data security is further ensured by setting different viewing permissions and multi-factor authentication. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solution of the present application, the following is a brief introduction to the drawings required for the description. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0016] Figure 1 This is a flow chart of the enterprise operating cost analysis method based on privacy computing of the present invention.

[0017] Figure 2 This is a schematic diagram of the enterprise operating cost analysis device based on privacy computing of the present invention. DETAILED DESCRIPTION

[0018] The specific steps of the enterprise operating cost analysis method based on privacy computing will be described in detail below, and various embodiments of the present disclosure will be described more comprehensively. The present disclosure can have various embodiments, and adjustments and changes can be made therein. However, it should be understood that there is no intention to limit the various embodiments of the present disclosure to the specific embodiments disclosed herein, but rather that the present disclosure should be understood to cover all adjustments, equivalents and / or alternatives that fall within the spirit and scope of the various embodiments of the present disclosure.

[0019] For example, as businesses continue to expand, operating cost analysis plays an increasingly important role in guiding corporate decision-making. Traditionally, this approach relies on integrating relevant data from various internal departments and external sources for comprehensive analysis. However, this approach carries data privacy risks, as data from departments like finance, sales, and production often contains sensitive information such as employee privacy and corporate trade secrets. Once this data is centralized for analysis, it poses a serious threat of data leakage.

[0020] Despite the availability of privacy-preserving technologies such as encryption, these technologies still face numerous challenges in their practical application in enterprise cost analysis. The implementation of some encryption technologies can significantly degrade computing performance, making the analysis process lengthy and complex, and unable to meet the needs of enterprises for real-time decision-making. Furthermore, during data interaction and collaborative computing, privacy controls are difficult to accurately manage, creating a constant risk of data leakage.

[0021] To address these issues, the industry has attempted to find solutions by optimizing encryption algorithms and establishing data access rights management. However, these efforts have not yielded the desired results. Optimized encryption algorithms still struggle to meet the complex and ever-changing computing requirements of enterprise operating cost analysis, while data rights management cannot effectively address cross-departmental and cross-institutional data collaboration scenarios. Therefore, the enterprise operating cost analysis field still lacks a solution that effectively protects data privacy while meeting the needs of efficient analysis.

[0022] To address the above issues, this embodiment provides an enterprise operating cost analysis method based on privacy computing. By integrating data from the enterprise's internal business system and external cooperation system, and utilizing federated learning and multi-party secure computing technologies, it achieves enterprise operating cost analysis while protecting data privacy.

[0023] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0024] See also Figure 1 FIG2 is a flowchart of a method for analyzing enterprise operating costs based on privacy computing in a specific embodiment, the method comprising the following steps: S1. Identify the enterprise's internal business systems and external collaborative systems as data sources, and set up data collection interfaces in each data source to collect, pre-process, and encrypt their respective business data; It should be noted that data sources are obtained from both inside and outside the enterprise, so that the enterprise operating cost analysis can be carried out based on comprehensive data; setting up data collection interfaces and performing collection, preprocessing and encryption operations ensure the quality and security of the data, preventing data from being leaked or damaged in the initial stage, and providing a data foundation for subsequent analysis; S2. Build a privacy-preserving computing platform. Each data source and the privacy-preserving computing platform use federated learning to construct a unified basic model for enterprise operating cost analysis. It should be noted that building a privacy-preserving computing platform provides a secure and efficient operating environment for federated learning and subsequent computational analysis. By building a unified basic model for enterprise operating cost analysis through federated learning, we fully leverage data from various data sources, integrate information from multiple departments or partners, improve the accuracy of the model, and better adapt to complex enterprise operating scenarios. S3. Split the data from each data source and send the split parts of the same data source to different data sources, thereby performing encrypted calculations between the data sources through multi-party secure computing, and uploading the encrypted calculation results to the privacy computing platform; It should be noted that data from the data source is segmented and encrypted through multi-party secure computing, ensuring the privacy of the data during the calculation process. Even if the data from a data source is leaked, the complete original data cannot be restored because the data has been segmented and the calculation process is encrypted, thus ensuring the security of enterprise data. At the same time, the integrated calculation of multi-party data provides data support for model optimization. S4. Use the encrypted calculation results on the privacy computing platform to optimize the unified basic model for enterprise operating cost analysis. Then, use the optimized unified basic model for enterprise operating cost analysis to perform enterprise operating privacy calculations, and set different viewing permissions for the enterprise operating privacy calculation results. It should be noted that the use of encrypted calculation results to optimize the basic model enables the model to accurately reflect the actual business operations of the enterprise and improves the accuracy of operating cost analysis; using the optimized model to perform business privacy calculations and setting different viewing permissions not only ensures the security of business privacy data, but also ensures that personnel with different permissions can obtain data related to it and in compliance with security regulations, so that enterprise data can be used and managed reasonably.

[0025] This embodiment provides a data foundation for subsequent calculations and analysis by determining the data source and performing data collection, preprocessing, and encryption operations. It builds a privacy computing platform and constructs a basic model through federated learning, enabling collaborative analysis of multiple data sources and improving model accuracy. The introduction of multi-party secure computing ensures the privacy protection of data during the calculation process, and the final model optimization, privacy computing, and permission setting provide enterprises with an operating cost analysis solution.

[0026] Furthermore, as a refinement and extension of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, another enterprise operating cost analysis method based on privacy computing is provided, which includes the following steps: S1. Identify the enterprise's internal business systems and external cooperative systems as data sources, and set up data collection interfaces in each data source to collect, pre-process, and encrypt their respective business data; the specific steps of step S1 are as follows: S11. Identify the business systems within the enterprise as internal data sources and the external collaborative systems as external data sources; For example, the enterprise's internal procurement system, production system, and financial system serve as internal data sources, and the enterprise's external supplier system and logistics supply system serve as external data sources. S12. Set up data collection interfaces for both internal and external data sources; S13 collects business data from the corresponding data source through the data acquisition interface, and pre-processes it in accordance with its own preset pre-processing method, and then performs encryption operations in accordance with the preset encryption method; For example, the product name, purchase price P are collected from the procurement system. p And the purchase quantity Q p ; Purchase unit price, use the median filling method; let the median of the purchase unit price be M p , then the missing value M p ; Use RSA encryption algorithm to encrypt, for example, use the encryption function: EncryptRSA(dt,key), where key is the encryption key; Collect production process P from the production system c , working hours H p , equipment depreciation De ; Standardize the working time data as follows:

[0027] in, is the mean working hours, is the standard deviation of working hours; it is encrypted using the AES encryption algorithm, and the encryption function is EncryptAES(dt,key); Collect expense category E from the financial system c Amount A e ; Logarithmic transformation is performed on the amount of expenditure to stabilize the variance, and the working hours are Alog=ln(A e +1), use the elliptic curve encryption algorithm ECC encryption, the encryption function is EncryptECC(dt,key); S2. Build a privacy-preserving computing platform. Each data source and the privacy-preserving computing platform use federated learning to build a unified basic model for enterprise operating cost analysis. The specific steps of step S2 are as follows: S21. Build a privacy-preserving computing platform and establish connections between the privacy-preserving computing platform and various data sources; Specifically, the privacy computing platform consists of multiple computing nodes, and a stable connection between the privacy computing platform and various data sources is established through the TCP / IP protocol; S22. Filter out the data source required for federated learning from the data source as the first data source, and deploy a federated learning client on the first data source; For example, the data sources of the procurement, production, and finance departments are selected as the first data source, and the federated learning client is deployed in each data source; S23. Each first data source constructs an enterprise operating cost analysis sub-model on its corresponding federated learning client and trains it using its own encrypted business data. After training, the sub-model parameters are obtained and encrypted and uploaded to the privacy computing platform. For example, a linear regression model is used in the federated learning client of the procurement system to build a procurement cost prediction sub-model:

[0028] in, Parameters of the procurement cost prediction sub-model, is the error term, P p is the purchase price, Q p is the purchase quantity; The least squares method is used to solve the parameters in the federated learning client of the procurement system, and the loss function is constructed as follows:

[0029] in, is the predicted value of procurement cost, is the actual value of purchase cost; Iteratively optimize the parameters of the procurement cost prediction sub-model based on the loss function; The federated learning client of the production system enables the neural network model to build a production cost prediction sub-model; The production cost prediction sub-model includes input layer, hidden layer and output layer. The input layer has three nodes: production process, standardized working hours and equipment depreciation. The hidden layer has ten nodes. The output layer is the production cost prediction value C. pd ; The activation function uses ReLU, the formula is ReLU ( x )=max(0, x ); trained by the back propagation algorithm, the loss function uses the mean square error function as follows: Where m is the number of product samples, is the production cost forecast, is the actual value of production cost; In the federated learning client of the financial system, a sub-model of the relationship between expenditure and cost is constructed using a decision tree algorithm; Specifically, a decision tree is constructed based on the expense category and amount characteristics, and the split attribute is selected through information gain. The information gain formula is as follows:

[0030] Among them, S is the sample set, A For attributes, V ( A ) is the attribute A The value set of S δ for A The value is δ The sample subset at time H is information entropy; After each department completes training, the model parameters are encrypted and uploaded to the privacy computing platform; S24. The privacy computing platform decrypts the parameters of each sub-model and performs weighted calculation according to the set weights to obtain the comprehensive weighted parameters. The comprehensive weighted parameters are returned to the federated learning clients of each first data source for retraining the sub-models until the iteration termination condition is met. The privacy computing platform decrypts the parameters of each sub-model and sets the procurement department model parameter as β pc , the production department model parameter is θ pd , the financial department model parameter is γ f ; According to the weight ω pc 、ωpd 、ω f Perform weighted calculation and obtain the comprehensive weighted parameters as follows: ; The comprehensive weighted parameters are returned to the federated learning client of each data source, and the sub-model training is re-performed until the iteration termination conditions are met, such as the loss function convergence or the preset number of iterations. S25. The federated learning client of each first data source returns each enterprise operating cost analysis sub-model to the privacy computing platform, thereby obtaining a unified enterprise operating cost analysis basic model. Specifically, the unified basic model for enterprise operating cost analysis is as follows: TC=a×C pc +b×C pd +c×C f Where TC is the total cost; C pc is the predicted value of procurement cost, which is calculated by the linear regression sub-model of the procurement system:

[0031] C pd is the production cost forecast, calculated by the neural network sub-model of the production system: C pd =NN(P c , , D e ) Among them, NN represents the output of the neural network model; C f is the predicted value of financial cost, which is calculated by the decision tree sub-model of the financial system: C f =DT(E c ,Alog) Among them, DT represents the output of the decision tree model; a, b, c are C pc 、C pd 、C f The weight coefficient of It should be noted that the unified basic model for enterprise operating cost analysis integrates sub-models of the procurement system, production system, and financial system, reflecting enterprise operating costs from multiple dimensions. For example, by inputting the purchase quantity and unit price, the procurement cost can be calculated through the procurement sub-model module; by inputting the production process and working hours, the production cost can be calculated through the production sub-model module; combined with the financial sub-model module, the impact of expense expenditure on total cost is comprehensively considered; S3. Split the data from each data source and send the split parts of the same data source to different data sources, thereby performing encrypted calculations between the data sources through multi-party secure computing, and uploading the encrypted calculation results to the privacy computing platform. The specific steps of step S3 are as follows: S31. Filter out the data source required for multi-party secure computing from the data source as the second data source, and determine the number N of the second data source; For example, the procurement system, production system, and financial system data sources are used as the second data source, N=3; S32. Split the data in each second data source to ensure that each record is split into N-1 parts; S33. Send the N-1 portion of data from each second data source to a different second data source to ensure that a complete data record of each second data source is not obtained by another second data source; For example, the purchase price and purchase quantity fields of each record in the purchase system are divided into two parts. For example, for the purchase price P p , send the odd-numbered rows of data to the production system, and the even-numbered rows of data to the financial system; for the purchase quantity Q p Send the even-numbered row data share to the production system, and the odd-numbered row data share to the financial system; Code the production process of the production system P c , standardized working hours The field is divided into two parts, and the production process code P c The odd-numbered rows of data are sent to the purchasing department, and the even-numbered rows of data are sent to the financial system; the standardized working hours are The even-numbered rows of data in a field are sent to the purchasing system, and the odd-numbered rows of data are sent to the financial system. Code the expense category of the financial system as E c The data Alog field after the amount logarithm transformation is split into two parts and sent to the procurement and production systems. Specifically, the expense category code E c The odd-numbered row data share is sent to the procurement system, and the even-numbered row data share is sent to the production system; the even-numbered row data share of the logarithmic transformed amount data Alog is sent to the procurement system, and the odd-numbered row data share is sent to the production system; S34. Each second data source calculates an intermediate result for its own partial data and the N-1 partial data obtained from each other second data source according to the logical relationship of the data fields. The intermediate result is encrypted, transmitted, and combined for calculation between the second data sources through the obfuscation circuit to obtain an encrypted calculation result. For example, after receiving the odd-numbered rows of purchase price data and the even-numbered rows of purchase quantity data from the purchasing department, the production system calculates intermediate results based on its own production data. For example, the intermediate result of the correlation between production material cost and production hours is calculated. The intermediate result is:

[0032] in, Purchase price odd-numbered row data, It is the even-numbered row data of the purchase quantity, and cof1 is the process coefficient determined based on production; The procurement system calculates the intermediate results of the relationship between procurement costs and expense expenditures based on the data received from the production department and the finance department:

[0033] in, is the partial purchase cost calculated based on its own data share, and cof2 is the correlation coefficient; The financial system generates the following intermediate results based on the received production department data and its own expenditure:

[0034] in, Is the received production process code, is the cost amount Ae related to the received production process code, is the depreciation of the equipment received, is the equipment depreciation rate; S35. Each second data source uploads its own encrypted calculation results to the privacy computing platform; S4. Optimize the unified enterprise operating cost analysis basic model using the encrypted calculation results on the privacy computing platform. Then, use the optimized unified enterprise operating cost analysis basic model to perform enterprise operating privacy calculations, and set different viewing permissions for the enterprise operating privacy calculation results. The specific steps of step S4 are as follows: S41. The privacy computing platform receives the encrypted calculation results uploaded by each second data source and decrypts them to obtain partial data from each data source; S42. The privacy computing platform uses partial data from various data sources to optimize the parameters of the unified enterprise operating cost analysis basic model until the model meets the requirements. Specifically, a unified basic model for enterprise operating cost analysis is obtained: TC=a×C pc +b×C pd +c×C f The privacy computing platform uses the decrypted data from the second data source and uses the stochastic gradient descent algorithm to adjust the weight coefficient. a '、 b '、 c ′, with the learning rate µ as an example, the gradient is calculated based on the data from the second data source:

[0035] in, is the loss function; The weight update formula is as follows: , ,

[0036] After multiple iterations, the model was optimized until it met the requirements. The unified basic model for enterprise operating cost analysis is as follows: TC= a ′×C pc + b ′×C pd + c ′×C f S43. The privacy computing platform obtains privacy computing requirements, parses the encrypted data from the corresponding business system as the data source, and inputs it into the optimized unified enterprise operating cost analysis basic model to complete the enterprise operation privacy computing. S44. The privacy computing platform will determine the privacy level of the enterprise's operational privacy calculation results and, based on the enterprise's internal user role permissions, assign corresponding viewing permissions to different user roles. S45. The privacy computing platform verifies the corresponding permissions of users who view the enterprise operation privacy computing results through multi-factor authentication.

[0037] In some embodiments, a distributed privacy computing platform is constructed in step S21, wherein the privacy computing platform includes a plurality of computing nodes; The privacy computing platform treats both comprehensive weighted parameter calculations and enterprise operation privacy calculations as computing tasks, and distributes the computing tasks to each computing node for parallel execution during the calculation process.

[0038] In some embodiments, the following steps are also included: The privacy computing platform records the time when each data source uploads model parameters and receives comprehensive weighted parameters, the sub-model training process of the federated learning client of each data source, the data segmentation in multi-party secure computing, and the calculation of intermediate results into the blockchain.

[0039] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0040] like Figure 2 As shown, the following is an embodiment of the enterprise operating cost analysis device based on privacy computing provided by the embodiment of the present disclosure. The device and the enterprise operating cost analysis method based on privacy computing in the above-mentioned embodiments belong to the same inventive concept. For details not fully described in the embodiment of the enterprise operating cost analysis device based on privacy computing, please refer to the embodiment of the above-mentioned enterprise operating cost analysis device method based on privacy computing.

[0041] The device includes: The data collection module is used to determine the enterprise's internal business systems and external cooperation systems as data sources. A data collection interface is set up in each data source to collect, pre-process and encrypt their respective business data; The privacy computing module is used to build a privacy computing platform. Various data sources and the privacy computing platform use federated learning to build a unified basic model for enterprise operating cost analysis. A multi-party secure computing module is used to split the data of each data source and send the split parts of the same data source to different data sources, thereby performing encrypted calculations between the data sources through multi-party secure computing and uploading the encrypted calculation results to the privacy computing platform; The enterprise operating cost analysis and calculation module is used to optimize the unified enterprise operating cost analysis basic model using encrypted calculation results on the privacy computing platform, and then use the optimized unified enterprise operating cost analysis basic model to perform enterprise operating privacy calculations, and set different viewing permissions for the enterprise operating privacy calculation results.

[0042] This embodiment realizes enterprise operating cost analysis through the collaborative work of the data collection module, privacy calculation module, multi-party secure calculation module and enterprise operating cost analysis calculation module, integrates complex processes, and improves the maintainability of the system.

[0043] The enterprise operating cost analysis method based on privacy computing provided in the embodiment of the present application can be applied to electronic devices. Those skilled in the art will understand that the electronic device structure involved in the embodiment of the present invention does not constitute a limitation on the electronic device, and the electronic device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently. In the embodiment of the present invention, the electronic device includes but is not limited to laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of the present application described and / or required herein.

[0044] The electronic device may include a processor, an external memory interface, an internal memory, a universal serial bus (USB) interface, a charging management module, a power management module, a battery, a wireless communication module, an audio module, a speaker, a microphone, a sensor module, a button, a camera, a display, and a SIM card interface, etc.

[0045] It is understood that the structures illustrated in the embodiments of the present application do not constitute specific limitations on the electronic device. In other embodiments of the present application, the electronic device may include more or fewer components than shown, or combine or separate certain components, or arrange the components differently. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0046] A processor may include one or more processing units, such as a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural network processing unit (NPU). Different processing units may be independent devices or integrated into one or more processors.

[0047] The processor can be the nerve center and command center of the electronic device. The controller can generate operation control signals based on the instruction opcode and timing signal to complete the control of instruction fetching and execution.

[0048] The processor may also include a memory for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. This memory can store instructions or data that the processor has just used or is reusing. If the processor needs to use the instruction or data again, it can directly call it from the memory. This avoids repeated accesses, reduces processor latency, and thus improves system efficiency.

[0049] The above-mentioned electronic device implements the enterprise operating cost analysis method based on privacy computing of this application, which determines the internal business system of the enterprise and the external cooperation system as data sources, and sets a data collection interface in each data source to collect, pre-process and encrypt their respective business data; builds a privacy computing platform, and each data source and the privacy computing platform build a unified enterprise operating cost analysis basic model through federated learning; divides the data of each data source, and sends the divided parts of the same data source to different data sources, so as to perform encryption calculations between each data source through multi-party secure computing, and upload the encrypted calculation results to the privacy computing platform; uses the encrypted calculation results on the privacy computing platform to optimize the unified enterprise operating cost analysis basic model, and then uses the optimized unified enterprise operating cost analysis basic model to perform enterprise business privacy calculation, and sets different viewing permissions for the enterprise business privacy calculation results, so as to achieve the beneficial effect of enterprise operating cost analysis under the premise of protecting data privacy by integrating the data of the internal business system and the external cooperation system of the enterprise and using federated learning and multi-party secure computing.

[0050] The storage medium provided in this application stores a program product that can implement an enterprise operating cost analysis method based on privacy computing.

[0051] The enterprise operating cost analysis method based on privacy computing includes: determining the internal business system of the enterprise and the external cooperation system as data sources, setting up a data collection interface in each data source to collect, pre-process and encrypt their respective business data; building a privacy computing platform, and each data source and the privacy computing platform build a unified enterprise operating cost analysis basic model through federated learning; dividing the data of each data source, and sending the divided parts of the same data source to different data sources, thereby performing encryption calculations between each data source through multi-party secure computing, and uploading the encrypted calculation results to the privacy computing platform; using the encrypted calculation results on the privacy computing platform to optimize the unified enterprise operating cost analysis basic model, and then using the optimized unified enterprise operating cost analysis basic model to perform enterprise operating privacy calculations, and setting different viewing permissions for the enterprise operating privacy calculation results.

[0052] In some possible implementations, the enterprise operating cost analysis method based on privacy computing disclosed herein can be implemented in the form of a program product, which includes program code. When the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps described in the above "Exemplary Method" section of this specification according to various exemplary implementations of the present disclosure.

[0053] The storage medium of the present disclosure 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 be, for example, but not limited to, a system, device or component of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0054] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

[0055] The enterprise operating cost analysis method based on privacy computing provided in the embodiment of the present application can be applied to electronic devices. Those skilled in the art will understand that the electronic device structure involved in the embodiment of the present invention does not constitute a limitation on the electronic device, and the electronic device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently. In the embodiment of the present invention, the electronic device includes but is not limited to laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of the present application described and / or required herein.

[0056] The electronic device may include a processor, an external memory interface, an internal memory, a universal serial bus (USB) interface, a charging management module, a power management module, a battery, a wireless communication module, an audio module, a speaker, a microphone, a sensor module, a button, a camera, a display, and a SIM card interface, etc.

[0057] It is understood that the structures illustrated in the embodiments of the present application do not constitute specific limitations on the electronic device. In other embodiments of the present application, the electronic device may include more or fewer components than shown, or combine or separate certain components, or arrange the components differently. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0058] A processor may include one or more processing units, such as a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural network processing unit (NPU). Different processing units may be independent devices or integrated into one or more processors.

[0059] The processor can be the nerve center and command center of the electronic device. The controller can generate operation control signals based on the instruction opcode and timing signal to complete the control of instruction fetching and execution.

[0060] The processor may also include a memory for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. This memory can store instructions or data that the processor has just used or is reusing. If the processor needs to use the instruction or data again, it can directly call it from the memory. This avoids repeated accesses, reduces processor latency, and thus improves system efficiency.

[0061] The above-mentioned electronic device implements the enterprise operating cost analysis method based on privacy computing of this application, which determines the internal business system of the enterprise and the external cooperation system as data sources, and sets a data collection interface in each data source to collect, pre-process and encrypt their respective business data; builds a privacy computing platform, and each data source and the privacy computing platform build a unified enterprise operating cost analysis basic model through federated learning; divides the data of each data source, and sends the divided parts of the same data source to different data sources, so as to perform encryption calculation between each data source through multi-party secure computing, and upload the encrypted calculation results to the privacy computing platform; uses the encrypted calculation results on the privacy computing platform to optimize the unified enterprise operating cost analysis basic model, and then uses the optimized unified enterprise operating cost analysis basic model to perform enterprise business privacy calculation, and sets different viewing permissions for the enterprise business privacy calculation results, so as to achieve the beneficial effect of enterprise operating cost analysis under the premise of protecting data privacy by integrating the data of the internal business system and the external cooperation system of the enterprise and utilizing federated learning and multi-party secure computing technologies.

[0062] The storage medium provided in this application stores a program product that can implement an enterprise operating cost analysis method based on privacy computing.

[0063] The enterprise operating cost analysis method based on privacy computing includes: determining the internal business system of the enterprise and the external cooperation system as data sources, setting up a data collection interface in each data source to collect, pre-process and encrypt their respective business data; building a privacy computing platform, and each data source and the privacy computing platform build a unified enterprise operating cost analysis basic model through federated learning; dividing the data of each data source, and sending the divided parts of the same data source to different data sources, thereby performing encryption calculations between each data source through multi-party secure computing, and uploading the encrypted calculation results to the privacy computing platform; using the encrypted calculation results on the privacy computing platform to optimize the unified enterprise operating cost analysis basic model, and then using the optimized unified enterprise operating cost analysis basic model to perform enterprise operating privacy calculations, and setting different viewing permissions for the enterprise operating privacy calculation results.

[0064] In some possible implementations, the enterprise operating cost analysis method based on privacy computing disclosed herein can be implemented in the form of a program product, which includes program code. When the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps described in the above "Exemplary Method" section of this specification according to various exemplary implementations of the present disclosure.

[0065] The storage medium of the present disclosure 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 be, for example, but not limited to, a system, device or component of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0066] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for analyzing enterprise operating costs based on privacy computing, characterized in that: The steps include: S1. Identify the enterprise's internal business systems and external collaborative systems as data sources, and set up data collection interfaces in each data source to collect, pre-process, and encrypt their respective business data; S2. Build a privacy-preserving computing platform. Each data source and the privacy-preserving computing platform use federated learning to construct a unified basic model for enterprise operating cost analysis. S3. Split the data from each data source and send the split parts of the same data source to different data sources, thereby performing encrypted calculations between the data sources through multi-party secure computing, and uploading the encrypted calculation results to the privacy computing platform; S4. Use the encrypted calculation results on the privacy computing platform to optimize the unified basic model for enterprise operating cost analysis, then use the optimized unified basic model for enterprise operating cost analysis to perform enterprise operating cost privacy calculations, and set different viewing permissions for the enterprise operating cost privacy calculation results.

2. The enterprise operating cost analysis method based on privacy computing according to claim 1 is characterized in that: The specific steps of step S1 are as follows: S11. Identify the business systems within the enterprise as internal data sources and the external collaborative systems as external data sources; S12. Set up data collection interfaces for both internal and external data sources; S13. The business data of the corresponding data source is collected through the data collection interface, and is pre-processed according to the respective preset pre-processing methods, and then encrypted according to the preset encryption method.

3. The enterprise operating cost analysis method based on privacy computing according to claim 2 is characterized in that: The specific steps of step S2 are as follows: S21. Build a privacy-preserving computing platform and establish connections between the privacy-preserving computing platform and various data sources; S22. Filter out the data source required for federated learning from the data source as the first data source, and deploy a federated learning client on the first data source; S23. Each first data source constructs an enterprise operating cost analysis sub-model on its corresponding federated learning client and trains it using its own encrypted business data. After training, the sub-model parameters are obtained and encrypted and uploaded to the privacy computing platform. S24. The privacy computing platform decrypts the parameters of each sub-model and performs weighted calculation according to the set weights to obtain the comprehensive weighted parameters. The comprehensive weighted parameters are returned to the federated learning clients of each first data source for retraining the sub-models until the iteration termination condition is met. S25. The federated learning clients of each first data source return each enterprise operating cost analysis sub-model to the privacy computing platform to obtain a unified enterprise operating cost analysis basic model.

4. The enterprise operating cost analysis method based on privacy computing according to claim 3 is characterized in that: The specific steps of step S3 are as follows: S31. Filter out the data source required for multi-party secure computing from the data source as the second data source, and determine the number N of the second data source; S32. Split the data in each second data source to ensure that each record is split into N-1 parts; S33. Send the N-1 portion of data from each second data source to a different second data source to ensure that a complete data record of each second data source is not obtained by another second data source; S34. Each second data source calculates an intermediate result for its own partial data and the N-1 partial data obtained from each other second data source according to the logical relationship of the data fields. The intermediate result is encrypted, transmitted, and combined for calculation between the second data sources through the obfuscation circuit to obtain an encrypted calculation result. S35. Each second data source uploads its own encrypted calculation result part to the privacy computing platform.

5. The enterprise operating cost analysis method based on privacy computing according to claim 4 is characterized in that: The specific steps of step S4 are as follows: S41. The privacy computing platform receives the encrypted calculation results uploaded by each second data source and decrypts them to obtain partial data from each data source; S42. The privacy computing platform uses partial data from various data sources to optimize the parameters of the unified enterprise operating cost analysis basic model until the model meets the requirements. S43. The privacy computing platform obtains privacy computing requirements, parses the encrypted data from the corresponding business system as the data source, and inputs it into the optimized unified enterprise operating cost analysis basic model to complete the privacy calculation of enterprise operating costs; S44. The privacy computing platform will determine the privacy level of the enterprise's operating cost privacy calculation results and, based on the internal user role permissions of the enterprise, assign corresponding viewing permissions to different user roles; S45. The privacy computing platform verifies the corresponding permissions of users who view the enterprise operation privacy computing results through multi-factor authentication.

6. The enterprise operating cost analysis method based on privacy computing according to claim 5 is characterized in that: In step S21, a distributed privacy computing platform is constructed, wherein the privacy computing platform includes a plurality of computing nodes; The privacy computing platform treats both comprehensive weighted parameter calculations and enterprise operation privacy calculations as computing tasks, and distributes the computing tasks to each computing node for parallel execution during the calculation process.

7. The enterprise operating cost analysis method based on privacy computing according to claim 6 is characterized in that: The following steps are also included: The privacy computing platform records the time when each data source uploads model parameters and receives comprehensive weighted parameters, the sub-model training process of the federated learning client of each data source, the data segmentation in multi-party secure computing, and the calculation of intermediate results into the blockchain.

8. An enterprise operating cost analysis device based on privacy computing, characterized in that: include: The data collection module is used to determine the enterprise's internal business systems and external cooperation systems as data sources. A data collection interface is set up in each data source to collect, pre-process and encrypt their respective business data; The privacy computing module is used to build a privacy computing platform. Various data sources and the privacy computing platform use federated learning to build a unified basic model for enterprise operating cost analysis. A multi-party secure computing module is used to split the data of each data source and send the split parts of the same data source to different data sources, thereby performing encrypted calculations between the data sources through multi-party secure computing and uploading the encrypted calculation results to the privacy computing platform; The enterprise operating cost analysis and calculation module is used to optimize the unified enterprise operating cost analysis basic model using encrypted calculation results on the privacy computing platform, and then use the optimized unified enterprise operating cost analysis basic model to perform enterprise operating cost privacy calculation, and set different viewing permissions for the enterprise operating cost privacy calculation results.

9. An electronic device, characterized in that: It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the enterprise operating cost analysis method based on privacy computing as described in any one of claims 1 to 7 are implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the enterprise operating cost analysis method based on privacy computing as described in any one of claims 1 to 7 are implemented.

Images