Privacy protection scheme optimization method, device and equipment based on consistency evaluation

Abstract

The invention provides a privacy protection scheme optimization method, device and equipment based on consistency evaluation, and belongs to the technical field of data processing.The method comprises the steps that protection capabilities of sensitive data of different types are processed in advance based on privacy protection algorithms of different mechanisms, and a first protection capability matrix is determined; determining a second protection capability matrix from the first protection capability matrix based on a plurality of privacy protection algorithms corresponding to the privacy protection scheme; determining a protection effect vector of the privacy protection scheme based on the second protection capability matrix, the protection demand vector and the influence mapping matrix; and based on the similarity between the protection effect vector and the business target vector, executing optimization operation of the privacy protection scheme. According to the method, the protection capability matrix is constructed in advance to map the protection capabilities of the privacy protection algorithms of different mechanisms to the unified space, the privacy demand, the privacy protection effect and the business target consistency are comprehensively considered, and the privacy protection scheme optimization is realized on the premise of not damaging the data availability.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular, to an optimization method, device, and equipment for a privacy protection scheme based on consistency evaluation. Background Art

[0002] With the wide application of communication technology and network technology, various types of sensitive data such as personal information and trade secrets are facing a greater risk of leakage or abuse.

[0003] Currently, in order to prevent the leakage or abuse of various types of sensitive data, the common method is to first process the sensitive data using a privacy protection scheme based on various privacy protection algorithms to obtain desensitized data, and then use the desensitized data for business processing.

[0004] However, after the sensitive data is desensitized using the privacy protection scheme, the privacy protection scheme often affects the usability of the data itself with respect to the business. Since only the impact of the privacy protection scheme on the privacy of the sensitive data is concerned, the situation where the desensitized data cannot be used for business may occur. Summary of the Invention

[0005] The present invention provides an optimization method, device, and equipment for a privacy protection scheme based on consistency evaluation, which is used to solve the defect in the prior art that the desensitized data cannot be used for business after the sensitive data is desensitized using the privacy protection scheme, and to implement a privacy protection scheme that does not affect the business usability of the sensitive data.

[0006] The present invention provides an optimization method for a privacy protection scheme based on consistency evaluation, including: Pre-determining a first protection capability matrix based on the protection capabilities of privacy protection algorithms of different mechanisms for processing different types of sensitive data; Determining a second protection capability matrix from the first protection capability matrix based on a plurality of privacy protection algorithms corresponding to the privacy protection scheme; Determining a protection effect vector of the privacy protection scheme based on the second protection capability matrix, a protection requirement vector, and an impact mapping matrix; the protection requirement vector is determined according to the sensitivity levels of different types of sensitive data; the impact mapping matrix is determined according to the contribution degrees of different types of historical sensitive data with respect to predefined business target metrics; Performing an optimization operation on the privacy protection scheme based on the similarity degree between the protection effect vector and a business target vector; the business target vector is pre-determined based on the importance of the business target metrics.

[0007] A method for optimizing a privacy protection scheme based on consistency evaluation according to the present invention includes: the privacy protection algorithms based on different mechanisms process the protection capabilities of different types of sensitive data, and a first protection capability matrix is determined in advance, including: Using the privacy protection algorithms of the different mechanisms to process the different types of sensitive data to obtain desensitized data corresponding to the different types of sensitive data; Based on the different types of sensitive data and their corresponding desensitized data, calculate the information loss values when the privacy protection algorithms of the different mechanisms process the different types of sensitive data; Normalize the information loss values to obtain the first protection capability matrix.

[0008] A method for optimizing a privacy protection scheme based on consistency evaluation according to the present invention includes: based on the second protection capability matrix, the protection requirement vector, and the impact mapping matrix, determining the protection effect vector of the privacy protection scheme, including: Based on the second protection capability matrix, the protection requirement vector, and the impact mapping matrix, calculate the privacy protection effect values when the several privacy protection algorithms process the different types of sensitive data to determine the effect evaluation matrix of the privacy protection scheme; Merge the privacy protection effect values when the several privacy protection algorithms in the effect evaluation matrix process the same type of sensitive data to obtain the merged effect vector of the privacy protection scheme; Normalize the merged effect vector to obtain the protection effect vector of the privacy protection scheme.

[0009] A method for optimizing a privacy protection scheme based on consistency evaluation according to the present invention includes: the protection requirement vector is determined based on the following method: Obtain a historical data set containing the different types of sensitive data; Perform clustering analysis on the historical data set to obtain the sensitive levels of the different types of sensitive data; Based on the sensitive levels of the different types of sensitive data, construct the protection requirement vector.

[0010] A method for optimizing a privacy protection scheme based on consistency evaluation according to the present invention includes: the business objective vector is determined based on the following method: Based on business requirements, pre-define the business objective indicators; Based on the double-base point method and the entropy weight method, analyze the importance among multiple business objective indicators to obtain the weights of the multiple business objective indicators; Construct the service objective vector based on the weights of the multiple service objective metrics.

[0011] According to a privacy protection scheme optimization method based on consistency evaluation provided by the present invention, it includes: performing an optimization operation on the privacy protection scheme based on the similarity between the protection effect vector and the service objective vector, including: When the similarity between the protection effect vector and the service objective vector reaches a preset similarity threshold, determine an optimal privacy protection scheme based on the privacy protection scheme; When the similarity between the protection effect vector and the service objective vector does not reach the preset similarity threshold, update the privacy protection scheme and / or the protection requirement vector, return to the step of determining a second protection ability matrix from the first protection ability matrix based on several privacy protection algorithms corresponding to the privacy protection scheme, obtain a new protection effect vector, and perform the optimization operation based on the similarity between the new protection effect vector and the service objective vector until the optimal privacy protection scheme is determined.

[0012] The present invention also provides a privacy protection scheme optimization device based on consistency evaluation, including: A first matrix determination module, configured to pre-determine a first protection ability matrix based on the protection capabilities of privacy protection algorithms of different mechanisms for processing different types of sensitive data; A second matrix determination module, configured to determine a second protection ability matrix from the first protection ability matrix based on several privacy protection algorithms corresponding to the privacy protection scheme; A protection effect determination module, configured to determine the protection effect vector of the privacy protection scheme based on the second protection ability matrix, the protection requirement vector, and the influence mapping matrix; the protection requirement vector is determined according to the sensitivity levels of different types of sensitive data; the influence mapping matrix is determined according to the contribution degrees of different types of historical sensitive data to predefined service objective metrics; A protection scheme optimization module, configured to perform an optimization operation on the privacy protection scheme based on the similarity between the protection effect vector and the service objective vector; the service objective vector is pre-determined based on the importance of the service objective metrics.

[0013] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and running on the processor, where when the processor executes the computer program, it implements the privacy protection scheme optimization method based on consistency evaluation as described in any one of the above.

[0014] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the privacy protection scheme optimization method based on consistency evaluation as described in any one of the above is implemented.

[0015] The present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, the privacy protection scheme optimization method based on consistency evaluation as described in any one of the above is implemented.

[0016] The privacy protection scheme optimization method, device and equipment based on consistency evaluation provided by the present invention determine a first protection capability matrix by pre-processing the protection capabilities of different types of sensitive data through privacy protection algorithms based on different mechanisms, so as to map the protection capabilities of cross-mechanism privacy protection algorithms or the same protection algorithm under different parameters to a unified space. After quantifying and comparing the effects of multiple privacy protection algorithms, a protection capability mapping mechanism of the privacy protection algorithm is constructed, which can effectively compare between privacy protection algorithms of different mechanisms, so as to realize the accurate evaluation and optimization of the privacy protection effect of a specific privacy protection scheme; by determining a second protection capability matrix from the first protection capability matrix according to a specific privacy protection scheme, and combining a protection requirement vector and an impact mapping matrix to determine a protection effect vector, the consistency between the privacy protection scheme and the business objective is achieved; comprehensively considering privacy requirements, privacy protection effects and business objective consistency, privacy security is ensured without compromising data availability, and the optimization of the privacy protection scheme is realized. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 is a flowchart of the privacy protection scheme optimization method based on consistency evaluation provided by the present invention.

[0019] Figure 2 is a structural diagram of the privacy protection scheme optimization device based on consistency evaluation provided by the present invention.

[0020] Figure 3 is a structural diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] To make the objectives, technical solutions and advantages of the present invention more clear, the following will, in conjunction with the accompanying drawings in the present invention, clearly and completely describe the technical solutions in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0022] It should be noted that in the description of the present invention, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, such that a process, method, article or device including a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0023] The terms "first", "second", etc. in the present invention are used to distinguish similar objects and are not used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first", "second", etc. are generally of the same category, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" means at least one of the connected objects, and the character " / " generally indicates an "or" relationship between the associated objects before and after.

[0024] The following will, in conjunction with Figures 1 - 3 describe the optimization method, device and equipment of the privacy protection scheme based on consistency evaluation provided by the present invention.

[0025] Figure 1 is a schematic flowchart of the optimization method of the privacy protection scheme based on consistency evaluation provided by the present invention. As Figure 1 shown, the method includes but is not limited to steps 101 to 104.

[0026] It should be noted that the execution subject of the optimization method of the privacy protection scheme based on consistency evaluation provided by the present invention is the corresponding optimization device of the privacy protection scheme based on consistency evaluation, which can specifically be a server, a computer device, such as a mobile phone, a tablet computer, a notebook computer, a handheld computer, a vehicle-mounted electronic device, a wearable device, an Ultra-Mobile Personal Computer (UMPC), a netbook or a Personal Digital Assistant (PDA), etc.

[0027] Step 101: Determine a first protection capability matrix in advance based on the protection capabilities of privacy protection algorithms with different mechanisms for different types of sensitive data.

[0028] Privacy protection algorithms with different mechanisms refer to different types of privacy protection algorithms with different implementation principles, different mathematical bases, and / or different parameters.

[0029] Among them, privacy protection algorithms with different mechanisms include, but are not limited to, different types of privacy protection algorithms such as anonymization algorithms and perturbation-based differential privacy algorithms; anonymization algorithms are applicable to protecting the overall pattern information of data, including, but not limited to, L-Diversity algorithm (l = 2), T-Closeness algorithm (t = 1.0), Topdown algorithm, etc.; perturbation-based differential privacy algorithms reduce the possibility for attackers to obtain accurate data through a noise addition mechanism, including, but not limited to, differential privacy algorithms with different noise addition mechanisms such as LaplaceBoundedDomain algorithm (δ = 0.0), LaplaceBoundedDomain algorithm (δ = 0.2), Laplace algorithm (δ = 0.2), etc.

[0030] Among them, the specific types of sensitive data can be determined according to specific business scenarios.

[0031] For example, in the business scenario of an electric power company optimizing power consumption scheduling, the specific types of sensitive data include, but are not limited to, sensitive data such as the power consumption history data of power users, the equipment installation location, and the personal information of power users.

[0032] Specifically, before using a privacy protection solution to process sensitive data, determine a first protection capability matrix in advance according to the protection capabilities of privacy protection algorithms with multiple different mechanisms for different types of sensitive data. At this time, each element value in the first protection capability matrix is the protection capability value of a privacy protection algorithm with one mechanism for one type of sensitive data.

[0033] Step 102: Determine a second protection capability matrix from the first protection capability matrix based on several privacy protection algorithms corresponding to the privacy protection solution.

[0034] The privacy protection solution actually applied to process different types of sensitive data consists of several privacy protection algorithms. The usage methods of different privacy protection algorithms are all regarded as different privacy protection solutions. There are different types, parameters, or combination orders, etc. among several privacy protection algorithms. That is, different privacy protection solutions include several types of different privacy protection algorithms, or include several privacy protection algorithms used in different combination orders, or include privacy protection algorithms with different parameters. The present invention does not limit this.

[0035] Specifically, when optimizing a privacy protection scheme for processing sensitive data, according to several privacy protection algorithms included in the privacy protection scheme, determine the protection ability values of the several privacy protection algorithms in the privacy protection scheme for processing different types of sensitive data from a pre-determined first protection ability matrix, so as to determine a second protection ability matrix specific to the privacy protection scheme.

[0036] It can be understood that when there is only one privacy protection algorithm corresponding to the privacy protection scheme, the second protection ability matrix is a row matrix or a column matrix.

[0037] Optionally, the privacy protection scheme includes at least one anonymization algorithm and at least one perturbation-based differential privacy algorithm, which can provide multi-level protection for data and further strengthen privacy protection.

[0038] Step 103: Based on the second protection ability matrix, the protection requirement vector, and the impact mapping matrix, determine the protection effect vector of the privacy protection scheme.

[0039] The protection requirement vector is determined according to the sensitivity levels of different types of sensitive data, and is used to represent the intensity requirements of different types of sensitive data in privacy protection.

[0040] The impact mapping matrix is determined according to the contribution degrees of different types of historical sensitive data with respect to predefined business target metrics.

[0041] The predefined business target metrics are target metrics determined based on business requirements in a certain business scenario.

[0042] Specifically, when optimizing a privacy protection scheme for processing sensitive data, in addition to determining a second protection ability matrix specific to the privacy protection scheme from a pre-determined first protection ability matrix, also determine the protection requirement vector according to the sensitivity levels of different types of sensitive data in the business scenario, and determine the impact mapping matrix according to the contribution degrees of different types of historical sensitive data with respect to predefined business target metrics. Further, according to the second protection ability matrix, the protection requirement vector, and the impact mapping matrix, calculate and determine the protection effect vector when the privacy protection scheme characterized by the second protection ability matrix processes different types of sensitive data characterized by the protection requirement vector, so as to obtain an index for evaluating the privacy protection effect of the privacy protection scheme.

[0043] Step 104: Based on the similarity degree between the protection effect vector and the business target vector, perform the optimization operation of the privacy protection scheme.

[0044] The business target vector is pre-determined based on the importance of the business target metrics.

[0045] Specifically, a business objective vector is determined in advance according to the importance among predefined business objective metrics, and then the correlation between the privacy protection effect and the business objective is determined by calculating the similarity between the protection effect vector and the business objective vector, so as to further perform optimization operations on the privacy protection scheme.

[0046] The method for optimizing a privacy protection scheme based on consistency evaluation provided by the present invention determines a first protection ability matrix by pre-determining the protection ability of privacy protection algorithms of different mechanisms for processing different types of sensitive data, thereby mapping the protection ability of cross-mechanism privacy protection algorithms or the same protection algorithm under different parameters to a unified space, quantifying and comparing the effects of multiple privacy protection algorithms, and constructing a protection ability mapping mechanism for privacy protection algorithms, which can effectively compare between privacy protection algorithms of different mechanisms, so as to achieve accurate evaluation and optimization of the privacy protection effect of a specific privacy protection scheme; by determining a second protection ability matrix from the first protection ability matrix according to the specific privacy protection scheme, and combining the protection requirement vector and the influence mapping matrix to determine the protection effect vector, the consistency between the privacy protection scheme and the business objective is achieved; comprehensively considering privacy requirements, privacy protection effects and business objective consistency, ensuring privacy security without compromising data availability, and realizing the optimization of the privacy protection scheme.

[0047] Based on the above embodiments, as an optional embodiment, the pre-determination of the first protection ability matrix according to the protection ability of privacy protection algorithms of different mechanisms for processing different types of sensitive data includes: Using the privacy protection algorithms of different mechanisms to process the different types of sensitive data to obtain the desensitized data corresponding to the different types of sensitive data; Based on the different types of sensitive data and their corresponding desensitized data, calculating the information loss value when the privacy protection algorithms of different mechanisms process the different types of sensitive data; Normalizing the information loss value to obtain the first protection ability matrix.

[0048] Specifically, when determining the first protection ability matrix according to the protection ability of privacy protection algorithms of different mechanisms for processing different types of sensitive data in advance, the privacy protection algorithms of different mechanisms are used to process different types of sensitive data to obtain the desensitized data corresponding to the different types of sensitive data.

[0049] For each privacy protection algorithm, regarding it as a black box, using the method in information theory and based on the information loss calculation index, calculating the information loss value between the information amount of different types of sensitive data and the information amount of its corresponding desensitized data as the information loss value caused by privacy protection when the privacy protection algorithms of different mechanisms act on different types of sensitive data.

[0050] Normalize the information loss values obtained after the privacy protection algorithms for all mechanisms process all types of sensitive data, and obtain the first protection capability matrix that establishes a mapping relationship for the privacy protection algorithms of all mechanisms. . Among them, represents the protection capability of the -th privacy protection algorithm for the -th type of sensitive data.

[0051] For example, in the case where the business scenario is the scenario of optimizing power consumption scheduling by an electric power company, the privacy protection algorithms of different mechanisms successively include the L-Diversity algorithm (l = 2), the T-Closeness algorithm (t = 1.0), the Topdown algorithm, the LaplaceBoundedDomain algorithm (δ = 0.0), the LaplaceBoundedDomain algorithm (δ = 0.2), and the Laplace algorithm (δ = 0.2), and the different types of sensitive data successively include the power consumption history data of power users, the equipment installation location, and the personal information of power users. After calculating the information loss values when the above six different mechanisms' privacy protection algorithms process the above three different types of sensitive data and normalizing them, the generated first protection capability matrix is as follows: ; Among them, each row in the first protection capability matrix represents the privacy protection algorithms of the above six different mechanisms, each column represents the above three different types of sensitive data, and each element value represents the protection capability after normalizing the information loss value.

[0052] For example, represents that the protection capability of the L-Diversity algorithm (l = 2) for the power consumption history data of power users is 0.75.

[0053] Optionally, calculating the information loss values when the privacy protection algorithms of the different mechanisms process the different types of sensitive data includes: determining the information loss values based on the information loss rate (Normalization of ConfidentialityProbability, NCP) or the generalized confidentiality probability (Generalized Confidentiality Probability, GCP). Among them, NCP is used to evaluate the information loss degree of a single equivalence class, and GCP is used to evaluate the information loss degree of the entire anonymized table.

[0054] The optimization method for privacy protection scheme based on consistency evaluation provided by the present invention evaluates the protection capabilities of privacy protection algorithms with different parameters and different types by calculating the information loss between sensitive data and desensitized data processed by privacy protection algorithms through different mechanisms and normalizing it, establishes the mapping relationship between the protection capabilities of different privacy protection algorithms, and maps the protection capabilities of privacy protection algorithms with different mechanisms such as different types of privacy protection algorithms, different privacy protection algorithms of the same type, and the same privacy protection algorithm under different parameter combinations to a unified space, so as to better understand and compare the performances of different privacy protection algorithms.

[0055] Based on the above embodiments, as an alternative embodiment, the protection requirement vector is determined in the following manner: Obtain a historical data set containing the different types of sensitive data; Perform clustering analysis on the historical data set to obtain the sensitivity levels of the different types of sensitive data; Construct the protection requirement vector based on the sensitivity levels of the different types of sensitive data.

[0056] Specifically, when constructing the protection requirement vector, obtain a historical data set containing different types of sensitive data, perform clustering analysis on the historical data set to obtain the sensitivity levels of the different types of sensitive data, and construct the protection requirement vector according to the sensitivity levels of the different types of sensitive data , denotes the privacy protection intensity requirement for the sensitive data of the th class. Among them, the number of elements in the protection requirement vector is the same as the number of types of sensitive data.

[0057] For example, in the case where the business scenario is the scenario of optimizing power consumption scheduling by an electric power company, the different types of sensitive data include three types of power data: the power consumption history data of power users, the equipment installation location, and the personal information of power users. Obtain the historical data set of these sensitive data, and use clustering analysis to perform clustering analysis on the historical data set to obtain that the sensitivity level of the power consumption history data of power users is high-sensitivity data, the sensitivity level of the equipment installation location is medium-sensitivity data, and the sensitivity level of the personal information of power users is low-sensitivity data.

[0058] Based on the sensitivity levels of the three types of power data, combined with historical privacy leakage events and user feedback, quantify the protection requirements for each type of data, and obtain that the protection requirement for high-sensitivity data (the power consumption history data of power users) is 0.8, the protection requirement for medium-sensitivity data (the equipment installation location) is 0.5, and the protection requirement for low-sensitivity data (the personal information of power users) is 0.3, so as to generate the protection requirement vector .

[0059] The privacy protection scheme optimization method based on consistency evaluation provided by the present invention classifies and grades the historical data sets of different types of sensitive data through cluster analysis, divides the data into several sensitive levels according to sensitivity, and combines other elements based on the sensitive levels to mine privacy protection requirements and quantify them to obtain a protection requirement vector representing the privacy protection intensity requirements of various types of sensitive data, so as to use the protection requirement vector to calculate the protection effect vector of the privacy protection scheme, and can perform more accurate effect evaluation on the privacy protection scheme.

[0060] Based on the above embodiments, as an alternative embodiment, the business objective vector is determined in the following manner: Based on business requirements, the business objective indicators are predefined in advance; Based on the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) and the entropy weight method, the importance among multiple business objective indicators is analyzed to obtain the weights of multiple business objective indicators; Based on the weights of multiple business objective indicators, the business objective vector is constructed.

[0061] The Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) is a method used to solve the evaluation and ranking problems of multiple indicators and multiple solutions. Specifically, in the process of scheme evaluation, according to the nature of indicators and data, a set of optimal indicator data is used as the virtual positive ideal solution, and a set of worst indicator data is used as the virtual negative ideal solution. The pros and cons of the evaluated scheme are judged by comparing the distances of the scheme points from the positive and negative ideal points.

[0062] The entropy weight method is a multi-index decision-making method. By calculating the information entropy of each evaluation index, the weights of each index are determined. The core idea is to determine the weights according to the variability or information content of the indicators. For indicators with large information content (i.e., indicators with large changes), they will have higher weights, while indicators with small information content (i.e., indicators with small changes) will have lower weights.

[0063] Specifically, when determining the business objective vector, the business requirements in the business scenario are collected and analyzed in advance, and multiple business objective indicators are defined. Then, according to TOPSIS and the entropy weight method, the importance among multiple business objective indicators is analyzed to obtain the weights of multiple business objective indicators. Then, the business objective vector is constructed according to the weights of the business objective indicators , where represents the importance weight of the th business objective indicator.

[0064] For example, in the case where the business scenario is the scenario of optimizing power consumption scheduling for an electric power company, the business requirements in the process of optimizing power consumption scheduling are collected, and three business objective indicators are predefined according to these business requirements, specifically including: the grid load balance degree (in percentage) for measuring the uniformity of power transmission, the power consumption cost reduction rate (in percentage) for representing the proportion of the saved power consumption cost in the total cost, and the prediction accuracy rate (in percentage) for measuring the accuracy of the energy prediction model.

[0065] The TOPSIS and entropy weight method are used to analyze the importance among the three business objective indicators. Expert scoring and normalization processing can also be combined to determine that the weight of the grid load balance degree is 0.4, the weight of the power consumption cost reduction rate is 0.35, and the weight of the prediction accuracy rate is 0.25. Thus, in the case of combining actual operation data and business requirements, a quantified business objective vector is obtained. 。

[0066] The privacy protection scheme optimization method based on consistency evaluation provided by the present invention defines business objective indicators by collecting business requirements in a business scenario, and analyzes the importance among business objective indicators by using a multi-attribute decision-making method based on the double-base point method and the entropy weight method without presetting weights, obtains the weights of business objective indicators to construct a quantified business objective vector, and uses the protection requirement vector to calculate the protection effect vector of the privacy protection scheme, which can ensure the consistency between the privacy protection scheme and the business objectives in the effect evaluation.

[0067] In an embodiment, the impact mapping matrix is determined based on the following method: collecting a historical data set including different types of sensitive data, determining the contribution degree of different types of historical sensitive data with respect to each predefined business objective indicator based on expert analysis evaluation and normalization processing, and determining the impact mapping matrix according to the contribution degree of different types of historical sensitive data. 。

[0068] For example, in the case where the business scenario is the scenario of optimizing power consumption scheduling for an electric power company, the three predefined business objective indicators are respectively the grid load balance degree, the power consumption cost reduction rate, and the prediction accuracy rate. Different types of sensitive data include the power consumption history data of power users, the equipment installation location, and the personal information of power users. The impact mapping matrix determined according to the contribution degree of different types of historical sensitive data with respect to the predefined business objective indicators is as follows: as follows: ; wherein, the impact mapping matrix Each row represents different types of sensitive data, and each column represents three business objective indicators. Taking the first row as an example, the first row corresponds to the electricity consumption history data of power users, and the three values respectively represent the contributions of the electricity consumption history data of power users to the load balancing degree (0.6), the electricity cost reduction rate (0.3), and the prediction accuracy rate (0.1).

[0069] Optionally, by adjusting the structure and values of the impact mapping matrix different scenarios and analysis requirements can be adapted.

[0070] Based on the above embodiments, as an alternative embodiment, determining the protection effect vector of the privacy protection scheme based on the second protection ability matrix, the protection requirement vector, and the impact mapping matrix includes:[[]]END] Based on the second protection ability matrix, the protection requirement vector, and the impact mapping matrix, calculate the privacy protection effect values when the several privacy protection algorithms process different types of sensitive data to determine the effect evaluation matrix of the privacy protection scheme; Merge the privacy protection effect values when the several privacy protection algorithms in the effect evaluation matrix process the same type of sensitive data to obtain the merged effect vector of the privacy protection scheme; Normalize the merged effect vector to obtain the protection effect vector of the privacy protection scheme.

[0071] Specifically, to evaluate the protection effect of a privacy protection scheme, it is necessary to determine the protection effect vector of the privacy protection scheme. Specifically, according to several privacy protection algorithms specifically used in the privacy protection scheme, determine the second protection ability matrix from the first protection ability matrix and determine the protection requirement vector according to the sensitivity levels of different types of sensitive data , determine the impact mapping matrix according to the contribution degrees of different types of historical sensitive data to predefined business objective indicators , and based on the second protection ability matrix , the protection requirement vector , and the impact mapping matrix , calculate the privacy protection effect values when each privacy protection algorithm processes different types of sensitive data, and construct the effect evaluation matrix of the privacy protection scheme according to the privacy protection effect values of all privacy protection algorithms in the privacy protection scheme processing all types of sensitive data . .

[0072] Further merge the privacy protection effect values when all privacy protection algorithms of the privacy protection scheme in the effect evaluation matrix process the same type of sensitive data to obtain the merged effect vector of the privacy protection scheme . ; Normalize the combined effect vector to obtain the protection effect vector of the privacy protection scheme . Among them, each element value in the protection effect vector reflects the protection effect of the privacy protection scheme on a certain business objective indicator.

[0073] Optionally, the expression for calculating the privacy protection effect value when the several privacy protection algorithms process different types of sensitive data based on the second protection ability matrix, the protection requirement vector, and the impact mapping matrix is as follows: ; Among them, is the privacy protection effect value in the th row and th column of the effect evaluation matrix , that is, the privacy protection effect of the th privacy protection algorithm in the privacy protection scheme on the th type of sensitive data; is the number of types of sensitive data; is the protection ability of the th privacy protection algorithm relative to the th type of sensitive data; is the protection requirement for the th type of sensitive data; is the contribution degree of the th type of sensitive data relative to the th business objective indicator.

[0074] Taking the privacy protection algorithms corresponding to the privacy protection scheme including the L-Diversity algorithm (l = 2) and the Laplace algorithm (δ = 0.2) as an example, the second protection ability matrix determined from the first protection ability matrix is as follows: .

[0075] In the case of the business scenario of optimizing power consumption scheduling by an electric power company, the effect evaluation matrix obtained based on the second protection ability matrix , the protection requirement vector and the impact mapping matrix is as follows: .

[0076] Merge the privacy protection effect values when the privacy protection algorithms in the effect evaluation matrix process the same type of sensitive data to obtain the combined effect vector of the privacy protection scheme.

[0077] Normalize the combined effect vector to obtain the protection effect vector of the privacy protection scheme .

[0078] The optimization method of the privacy protection scheme based on consistency evaluation provided by the present invention generates the protection effect vector of the privacy protection scheme by combining the second protection ability matrix, the protection requirement vector and the influence mapping matrix, and quantitatively evaluates and analyzes the privacy protection effect of the privacy protection scheme on each business objective index by fully considering the balance among the business objectives, the privacy protection requirements and the privacy protection effects. Furthermore, when optimizing the privacy protection scheme based on the protection effect vector, it can not only protect the user privacy, but also ensure the effective realization of the business objectives.

[0079] Based on the above embodiments, as an alternative embodiment, performing the optimization operation of the privacy protection scheme based on the similarity between the protection effect vector and the business objective vector includes: When the similarity between the protection effect vector and the business objective vector reaches a preset similarity threshold, determining the optimal privacy protection scheme based on the privacy protection scheme; When the similarity between the protection effect vector and the business objective vector does not reach the preset similarity threshold, updating the privacy protection scheme and / or the protection requirement vector, returning to the step of determining the second protection ability matrix from the first protection ability matrix based on the corresponding privacy protection algorithms of the privacy protection scheme, obtaining a new protection effect vector, and performing the optimization operation based on the similarity between the new protection effect vector and the business objective vector until the optimal privacy protection scheme is determined.

[0080] Optionally, the similarity between the protection effect vector and the business objective vector is determined based on methods such as Manhattan distance, Pearson correlation coefficient, Jaccard similarity, etc.

[0081] Optionally, the similarity between the protection effect vector and the business objective vector is determined based on the following expression: ; where is the included angle between the protection effect vector and the business objective vector , which is used to characterize the similarity between the protection effect vector and the business objective vector. The larger the included angle, the lower the similarity; is the dot product of the protection effect vector and the business objective vector ; is the norm of the protection effect vector ; is the norm of the business objective vector ;

[0082] Specifically, when performing the optimization operation of the privacy protection scheme, a consistency check is performed on the protection effect vector and the business objective vector, that is, the similarity degree between the protection effect vector and the business objective vector is calculated. The greater the similarity degree, the stronger the correlation between the privacy protection effect and the business objective, and the better the balance between privacy requirements and business effects can be achieved.

[0083] When the similarity degree reaches the preset similarity degree threshold, the direction consistency between the protection effect vector and the business objective vector is high, indicating that the privacy protection scheme is reasonable and meets the business requirements. There is no need to optimize, and the existing privacy protection scheme can be directly used as the optimal privacy protection scheme.

[0084] When the similarity degree does not reach the preset similarity degree threshold, it indicates that there is a significant deviation between the privacy protection effect distribution of the privacy protection scheme and the business objective requirement distribution, and the business availability of the desensitized data is poor. It is necessary to optimize the privacy protection scheme with the goal of increasing the similarity degree (such as reducing the included angle, etc.), specifically including updating the privacy protection scheme and / or the protection requirement vector.

[0085] Among them, the ways to update the privacy protection scheme include but are not limited to adjusting the number and / or combination order of privacy protection algorithms in the privacy protection scheme, and adjusting the parameters of privacy protection algorithms in the privacy protection scheme (such as increasing the k value of the anonymization algorithm and reducing the value) of differential privacy.

[0086] Return to a number of privacy protection algorithms corresponding to the privacy protection scheme based on the updated privacy protection scheme and / or protection requirement vector, determine the second protection ability matrix from the first protection ability matrix, re-obtain the new protection effect vector, re-calculate the similarity degree between the new protection effect vector and the business objective vector, and perform the optimization operation of the privacy protection scheme according to the similarity degree until the similarity degree between the new protection effect vector and the business objective vector reaches the preset similarity degree threshold, and the directions of the new protection effect vector and the business objective vector are close enough to determine the optimal privacy protection scheme.

[0087] The privacy protection scheme optimization method based on consistency evaluation provided by the present invention performs consistency analysis on the protection effect vector and the business objective vector by calculating the similarity degree of vectors, obtains a consistency score value reflecting the degree of consistency between the privacy protection effect and the business objective in the overall trend, and based on this score value, feedback-optimizes the privacy protection scheme, updates the number, order combination, parameters, and / or protection requirement vectors of the privacy protection algorithms in the privacy protection scheme, and re-performs a new round of evaluation and consistency test of the privacy protection scheme until the consistency test result reaches the expected standard, and finally forms an optimized privacy protection scheme.

[0088] Generally speaking, the privacy protection scheme optimization method based on consistency evaluation provided by the present invention obtains a business objective vector through quantitative analysis of the business objectives of the business scenario, obtains a protection requirement vector through analysis of the protection requirements of different types of sensitive data in the business scenario, determines a first protection ability matrix in advance based on the protection ability of privacy protection algorithms of different mechanisms to process different types of sensitive data through algorithm protection ability mapping analysis, determines a second protection ability matrix from the first protection ability matrix according to the privacy protection algorithms in the specific privacy protection scheme, combines the second protection ability matrix, the protection requirement vector, and the impact mapping matrix to determine a protection effect vector for evaluating the protection effect of the privacy protection scheme, calculates the similarity degree between the protection effect vector and the business objective vector for consistency test, and iteratively optimizes the privacy protection scheme according to the consistency test feedback.

[0089] The present invention can not only solve the problems of insufficient response to different privacy protection requirements and lack of consistency between privacy protection measures and business objectives. By mapping the protection ability of cross-mechanism privacy protection algorithms or the same protection algorithm under different parameters to a unified space based on information loss, quantifying and comparing the effects of multiple privacy protection algorithms, and constructing a protection ability mapping mechanism for privacy protection algorithms, it realizes the accurate evaluation and optimization of the privacy protection scheme effect. At the same time, through the feedback iteration mechanism, the present invention adjusts the scheme parameters or algorithm combination according to the consistency verification result, thereby realizing the dynamic balance and collaborative optimization of privacy protection and business objectives, and realizing the optimization of the privacy protection scheme considering privacy requirements, privacy protection effect, and business objective consistency.

[0090] Figure 2 It is a schematic structural diagram of the privacy protection scheme optimization device based on consistency evaluation provided by the present invention, as Figure 2 shown, the privacy protection scheme optimization device based on consistency evaluation includes but is not limited to a first matrix determination module 201, a second matrix determination module 202, a protection effect determination module 203, and a protection scheme optimization module 204.

[0091] The first matrix determination module 201 is configured to pre-determine a first protection capability matrix based on the protection capabilities of privacy protection algorithms with different mechanisms for processing different types of sensitive data.

[0092] The second matrix determination module 202 is configured to determine a second protection capability matrix from the first protection capability matrix based on a plurality of privacy protection algorithms corresponding to a privacy protection scheme.

[0093] The protection effect determination module 203 is configured to determine a protection effect vector of the privacy protection scheme based on the second protection capability matrix, a protection requirement vector, and an impact mapping matrix; the protection requirement vector is determined according to the sensitivity levels of different types of sensitive data; the impact mapping matrix is determined according to the contribution degrees of different types of historical sensitive data with respect to predefined service target metrics.

[0094] The protection scheme optimization module 204 is configured to perform an optimization operation on the privacy protection scheme based on the similarity degree between the protection effect vector and a service target vector; the service target vector is pre-determined based on the importance of the service target metrics.

[0095] It should be noted that the privacy protection scheme optimization device based on consistency evaluation provided by the present invention, during specific operation, can execute the privacy protection scheme optimization method described in any of the above embodiments, and this embodiment will not be elaborated herein.

[0096] The privacy protection scheme optimization device based on consistency evaluation provided by the present invention pre-determines a first protection capability matrix based on the protection capabilities of privacy protection algorithms with different mechanisms for processing different types of sensitive data, thereby mapping the protection capabilities of cross-mechanism privacy protection algorithms or the same protection algorithm under different parameters to a unified space, quantifying and comparing the effects of multiple privacy protection algorithms, and constructing a protection capability mapping mechanism for privacy protection algorithms, which can effectively compare between privacy protection algorithms with different mechanisms, so as to achieve accurate evaluation and optimization of the privacy protection effect of a specific privacy protection scheme; by determining a second protection capability matrix from the first protection capability matrix according to a specific privacy protection scheme, and combining the protection requirement vector and the impact mapping matrix to determine the protection effect vector, the consistency between the privacy protection scheme and the service target is achieved; comprehensively considering privacy requirements, privacy protection effects, and service target consistency, ensuring privacy security without compromising data availability, and realizing the optimization of the privacy protection scheme.

[0097] In one embodiment, the first matrix determination module is further configured to process the different types of sensitive data by using privacy protection algorithms of different mechanisms to obtain desensitized data corresponding to the different types of sensitive data; calculate information loss values when the privacy protection algorithms of different mechanisms process the different types of sensitive data based on the different types of sensitive data and their corresponding desensitized data; and normalize the information loss values to obtain the first protection ability matrix.

[0098] In one embodiment, the protection effect determination module is further configured to calculate privacy protection effect values when the plurality of privacy protection algorithms process the different types of sensitive data based on the second protection ability matrix, the protection requirement vector, and the influence mapping matrix to determine an effect evaluation matrix of the privacy protection scheme; merge the privacy protection effect values when the plurality of privacy protection algorithms in the effect evaluation matrix process the same type of sensitive data to obtain a merged effect vector of the privacy protection scheme; and normalize the merged effect vector to obtain a protection effect vector of the privacy protection scheme.

[0099] In one embodiment, the privacy protection scheme optimization device based on consistency evaluation further includes: a requirement vector determination module, configured to obtain a historical data set including the different types of sensitive data; perform clustering analysis on the historical data set to obtain the sensitive levels of the different types of sensitive data; and construct the protection requirement vector based on the sensitive levels of the different types of sensitive data.

[0100] In one embodiment, the privacy protection scheme optimization device based on consistency evaluation further includes: a service vector determination module, configured to predefined the service target indicators based on service requirements; analyze the importance among a plurality of the service target indicators by using the double base point method and the entropy weight method to obtain weights of the plurality of service target indicators; and construct the service target vector based on the weights of the plurality of service target indicators.

[0101] In one embodiment, the protection scheme optimization module is further configured to determine an optimal privacy protection scheme based on the determined privacy protection scheme when the similarity between the protection effect vector and the service target vector reaches a preset similarity threshold; when the similarity between the protection effect vector and the service target vector does not reach the preset similarity threshold, update the privacy protection scheme and / or the protection requirement vector, return to the step of determining the second protection ability matrix from the first protection ability matrix based on several privacy protection algorithms corresponding to the privacy protection scheme, obtain a new protection effect vector, and perform the optimization operation based on the similarity between the new protection effect vector and the service target vector until the optimal privacy protection scheme is determined.

[0102] Figure 3 is a schematic structural diagram of an electronic device provided by the present invention, as Figure 3 shown, the electronic device may include: a processor (Processor) 310, a communication interface (Communications Interface) 320, a memory (Memory) 330, and a communication bus 340. Among them, the processor 310, the communication interface 320, and the memory 330 communicate with each other through the communication bus 340. The processor 310 can call logical instructions in the memory 330 to execute the privacy protection scheme optimization method based on consistency evaluation provided in any of the above embodiments. The privacy protection scheme optimization method based on consistency evaluation includes but is not limited to the following steps: determining a first protection ability matrix in advance based on the protection capabilities of privacy protection algorithms of different mechanisms for processing different types of sensitive data; determining a second protection ability matrix from the first protection ability matrix based on several privacy protection algorithms corresponding to the privacy protection scheme; determining a protection effect vector of the privacy protection scheme based on the second protection ability matrix, the protection requirement vector, and the influence mapping matrix; the protection requirement vector is determined according to the sensitivity levels of different types of sensitive data; the influence mapping matrix is determined according to the contribution degrees of different types of historical sensitive data to predefined service target indicators; performing an optimization operation on the privacy protection scheme based on the similarity between the protection effect vector and the service target vector; the service target vector is determined in advance based on the importance of the service target indicators.

[0103] In addition, when the logical instructions in the above-mentioned memory 330 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0104] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the privacy protection scheme optimization method based on consistency evaluation provided in any of the above embodiments. The privacy protection scheme optimization method based on consistency evaluation includes, but is not limited to, the following steps: determining a first protection capability matrix in advance based on the protection capabilities of different privacy protection algorithms for different types of sensitive data; determining a second protection capability matrix from the first protection capability matrix based on several privacy protection algorithms corresponding to the privacy protection scheme; determining a protection effect vector of the privacy protection scheme based on the second protection capability matrix, a protection requirement vector, and an impact mapping matrix; the protection requirement vector is determined according to the sensitivity levels of different types of sensitive data; the impact mapping matrix is determined according to the contribution degrees of different types of historical sensitive data to predefined business target metrics; performing an optimization operation of the privacy protection scheme based on the similarity degree between the protection effect vector and a business target vector; the business target vector is determined in advance based on the importance of the business target metrics.

[0105] In another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the privacy protection scheme optimization method based on consistency evaluation provided by any one of the above embodiments. The privacy protection scheme optimization method based on consistency evaluation includes but is not limited to the following steps: determining a first protection capability matrix in advance based on the protection capabilities of privacy protection algorithms of different mechanisms for processing different types of sensitive data; determining a second protection capability matrix from the first protection capability matrix based on a plurality of privacy protection algorithms corresponding to the privacy protection scheme; determining a protection effect vector of the privacy protection scheme based on the second protection capability matrix, a protection requirement vector, and an influence mapping matrix; the protection requirement vector is determined according to the sensitivity levels of different types of sensitive data; the influence mapping matrix is determined according to the contribution degrees of different types of historical sensitive data to predefined business target metrics; performing an optimization operation on the privacy protection scheme based on the similarity degree between the protection effect vector and a business target vector; the business target vector is determined in advance based on the importance of the business target metrics.

[0106] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative effort.

[0107] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0108] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A privacy protection scheme optimization method based on consistency assessment, characterized in that: include: A first protection capability matrix is ​​predetermined based on the protection capabilities of privacy protection algorithms of different mechanisms to process different types of sensitive data; Based on a plurality of privacy protection algorithms corresponding to the privacy protection scheme, determining a second protection capability matrix from the first protection capability matrix; Determining a protection effect vector of the privacy protection scheme based on the second protection capability matrix, the protection requirement vector and the impact mapping matrix; the protection requirement vector is determined according to the sensitivity level of different types of sensitive data; the impact mapping matrix is ​​determined according to the contribution of different types of historical sensitive data to predefined business target indicators; Based on the similarity between the protection effect vector and the business goal vector, an optimization operation of the privacy protection scheme is performed; the business goal vector is predetermined based on the importance of the business goal indicator.

2. The privacy protection scheme optimization method based on consistency assessment according to claim 1 is characterized in that: The privacy protection algorithms based on different mechanisms process different types of sensitive data protection capabilities, and predetermine a first protection capability matrix, including: Processing the different types of sensitive data using the privacy protection algorithms of the different mechanisms to obtain desensitized data corresponding to the different types of sensitive data; Based on the different types of sensitive data and the desensitized data corresponding thereto, calculating information loss values ​​when the privacy protection algorithms of the different mechanisms process the different types of sensitive data; The information loss value is normalized to obtain the first protection capability matrix.

3. The privacy protection scheme optimization method based on consistency assessment according to claim 1 is characterized in that: The determining, based on the second protection capability matrix, the protection requirement vector and the impact mapping matrix, a protection effect vector of the privacy protection scheme includes: Based on the second protection capability matrix, the protection requirement vector and the impact mapping matrix, calculating the privacy protection effect values ​​when the plurality of privacy protection algorithms process the different types of sensitive data, so as to determine the effect evaluation matrix of the privacy protection scheme; Merging the privacy protection effect values ​​of the plurality of privacy protection algorithms in the effect evaluation matrix when processing the same type of sensitive data to obtain a combined effect vector of the privacy protection scheme; The combined effect vector is normalized to obtain a protection effect vector of the privacy protection scheme.

4. The privacy protection scheme optimization method based on consistency assessment according to claim 1 is characterized in that: The protection requirement vector is determined based on the following method: Obtaining a historical data set containing the different types of sensitive data; Performing cluster analysis on the historical data set to obtain the sensitivity levels of the different types of sensitive data; The protection requirement vector is constructed based on the sensitivity levels of the different types of sensitive data.

5. The privacy protection scheme optimization method based on consistency assessment according to claim 1 is characterized in that: The business goal vector is determined based on the following method: Based on business needs, pre-define the business target indicators; Analyze the importance of the multiple business objective indicators based on the double base point method and the entropy weight method to obtain the weights of the multiple business objective indicators; The business goal vector is constructed based on the weights of the plurality of business goal indicators.

6. The privacy protection scheme optimization method based on consistency assessment according to claim 1 is characterized in that: The performing the optimization operation of the privacy protection scheme based on the similarity between the protection effect vector and the business target vector includes: When the similarity between the protection effect vector and the business target vector reaches a preset similarity threshold, determining an optimal privacy protection scheme based on the privacy protection scheme; When the similarity between the protection effect vector and the business target vector does not reach the preset similarity threshold, the privacy protection scheme and / or the protection requirement vector are updated, and the process returns to the step of determining the second protection capability matrix from the first protection capability matrix based on the multiple privacy protection algorithms corresponding to the privacy protection scheme to obtain a new protection effect vector, and the optimization operation is performed based on the similarity between the new protection effect vector and the business target vector until the optimal privacy protection scheme is determined.

7. A privacy protection scheme optimization device based on consistency assessment, characterized in that: include: A first matrix determination module is used to determine the protection capabilities of different types of sensitive data based on privacy protection algorithms of different mechanisms, and to predetermine a first protection capability matrix; A second matrix determination module, configured to determine a second protection capability matrix from the first protection capability matrix based on a plurality of privacy protection algorithms corresponding to the privacy protection scheme; a protection effect determination module, configured to determine a protection effect vector of the privacy protection scheme based on the second protection capability matrix, a protection requirement vector and an impact mapping matrix; the protection requirement vector is determined according to the sensitivity levels of different types of sensitive data; the impact mapping matrix is ​​determined according to the contribution of different types of historical sensitive data to predefined business target indicators; A protection scheme optimization module is used to perform optimization operations of the privacy protection scheme based on the similarity between the protection effect vector and the business goal vector; the business goal vector is predetermined based on the importance of the business goal indicator.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the method for optimizing a privacy protection scheme based on consistency assessment as described in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for optimizing a privacy protection scheme based on consistency assessment as described in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for optimizing a privacy protection scheme based on consistency assessment as described in any one of claims 1 to 6 is implemented.

Images