Data sensitivity determination method, electronic device and computer-readable storage medium

Abstract

The present application discloses a data sensitivity determination method, an electronic device and a computer-readable storage medium, and relates to the field of data detection. By obtaining the initial data to be determined, identifying the initial data, obtaining multiple target fields and corresponding field weights of the initial data, obtaining preset sensitive categories according to the multiple target fields, and generating an initial matrix according to the multiple target fields and multiple field weights; calculating and obtaining a scoring matrix of the initial data according to the obtained preset sensitive categories, the preset weight database and the initial matrix, judging the sensitivity of the initial data according to the scoring matrix, it is possible to comprehensively analyze the sensitive information, improve the accuracy of identifying sensitive information, and effectively avoid the problem of information leakage.

Description

Technical Field

[0001] The present application relates to the field of data detection, and in particular to a data sensitivity determination method, an electronic device, and a computer-readable storage medium. Background Art

[0002] As the process of informatization gradually penetrates into daily life, digital storage of personal information, commercial secrets, and state secrets has become the mainstream, and the harm caused by various information leakage incidents has become more serious.

[0003] There are many main factors that cause data leakage at present. Taking the Internet industry, which accounts for the largest proportion, as an example, they include technical factors and non-technical factors. Among them, non-technical factors mainly include: internal employee leaks, poor internal information management, and human information leaks.

[0004] In order to strengthen the prevention of data leakage caused by non-technical reasons, the general solution is to implement sensitive data classification, information access permission control, information desensitization, encryption, watermarking, etc. Among them, taking sensitive data intelligent identification and classification technology as an example, only single data of sensitive information is analyzed, and a comprehensive analysis cannot be performed, resulting in low accuracy in sensitive information identification and failure to effectively prevent information leakage. Summary of the invention

[0005] The present application aims to solve at least one of the technical problems existing in the prior art. To this end, the present application proposes a data sensitivity determination method that can comprehensively analyze sensitive information, improve the accuracy of identifying sensitive information, and effectively avoid the problem of information leakage.

[0006] The present application also proposes an electronic device having the above-mentioned data sensitivity determination method.

[0007] The present application also proposes a computer-readable storage medium having the above-mentioned data sensitivity determination method.

[0008] According to the first aspect of the present application, the data sensitivity determination method includes:

[0009] Obtaining initial data to be determined;

[0010] Identify the initial data to obtain multiple target fields of the initial data and field weights corresponding to the target fields;

[0011] Acquire preset sensitive categories according to the multiple target fields, and generate an initial matrix according to the multiple target fields and multiple field weights;

[0012] Calculating a scoring matrix for the initial data according to the preset sensitive categories, the preset weight database and the initial matrix;

[0013] The sensitivity of the initial data is determined according to the scoring matrix. According to the data sensitivity determination method of the embodiment of the present application, it has at least the following beneficial effects: by obtaining the initial data to be determined, identifying the initial data, obtaining multiple target fields and corresponding field weights of the initial data, obtaining preset sensitive categories according to the multiple target fields and generating an initial matrix according to the multiple target fields and the multiple field weights; according to the obtained preset sensitive categories, the preset weight database and the initial matrix, the scoring matrix of the initial data is calculated, and the sensitivity of the initial data is determined according to the scoring matrix, which can comprehensively analyze the sensitive information, improve the accuracy of identifying sensitive information, and effectively avoid the problem of information leakage.

[0014] According to some embodiments of the present application, the initial data includes simple data;

[0015] The identifying the initial data to obtain a target field of the initial data and a field weight corresponding to the target field includes:

[0016] The simple data is identified based on field features of multiple target fields and regular expressions to obtain multiple target fields of the simple data and field weights corresponding to the target fields.

[0017] According to some embodiments of the present application, the initial data includes complex data;

[0018] The identifying the initial data to obtain a target field of the initial data and a field weight corresponding to the target field includes:

[0019] The complex data is identified based on a machine learning classification algorithm to obtain multiple target fields of the complex data and field weights corresponding to the target fields.

[0020] According to some embodiments of the present application, the weight database includes a plurality of verification weight matrices;

[0021] Correspondingly, the step of generating a scoring matrix for the initial data according to the preset sensitive category, the preset weight database and the initial matrix includes:

[0022] Acquire multiple verification weight matrices according to the preset sensitive categories and the preset weight database;

[0023] A scoring matrix for the initial data is calculated based on the multiple verification weight matrices and the initial matrix.

[0024] According to some embodiments of the present application, the weight database includes a plurality of verification weight matrices and a preset weight matrix;

[0025] Correspondingly, the step of generating a scoring matrix for the initial data according to the preset sensitive category, the preset weight database and the initial matrix includes:

[0026] Acquire a plurality of verification weight matrices and a preset weight matrix according to the preset sensitive category and the preset weight database;

[0027] A scoring matrix for the initial data is calculated based on the multiple verification weight matrices, the preset weight matrix and the initial matrix.

[0028] According to some embodiments of the present application, it also includes: adjusting the weight of the preset weight matrix according to the scoring matrix.

[0029] According to some embodiments of the present application, weight adjustment of the preset weight matrix according to the scoring matrix further includes:

[0030] Setting an expectation matrix according to the initial data;

[0031] Acquire an error matrix based on the expected matrix and the scoring matrix;

[0032] Calculate a vector matrix based on the error matrix and the preset weight matrix;

[0033] An adjusted preset weight matrix is ​​calculated based on the multiplication factor matrix and the scoring matrix.

[0034] According to some embodiments of the present application, determining the sensitivity of the initial data according to the scoring matrix includes:

[0035] Determining the optimal sensitive category corresponding to the initial data according to the scoring matrix;

[0036] The sensitivity of the initial data is determined according to the optimal sensitive category and a preset sensitive category database.

[0037] According to an electronic device of an embodiment of the second aspect of the present application, the device comprises: at least one processor, and a memory communicatively connected to the at least one processor; wherein the memory stores instructions, and the instructions are executed by the at least one processor so that the at least one processor implements the data sensitivity determination method as described in the first aspect when executing the instructions.

[0038] The electronic device according to the present application has at least the following beneficial effects: by executing the data sensitivity determination method mentioned in the embodiment of the first aspect, it is possible to comprehensively analyze sensitive information, improve the accuracy of identifying sensitive information, and effectively avoid the problem of information leakage.

[0039] According to the computer-readable storage medium of the third aspect embodiment of the present application, the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the data sensitivity determination method as described in the first aspect.

[0040] The computer-readable storage medium according to the present application has at least the following beneficial effects: by executing the data sensitivity determination method mentioned in the embodiment of the first aspect, it is possible to comprehensively analyze sensitive information, improve the accuracy of identifying sensitive information, and effectively avoid the problem of information leakage.

[0041] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 A schematic diagram of a specific process of the data sensitivity determination method in an embodiment of the present application;

[0043] Figure 2 This is a specific application example diagram of a preset sensitive category in an embodiment of the present application;

[0044] Figure 3 A specific flow chart of step S400 of the data sensitivity determination method in the embodiment of the present application;

[0045] Figure 4 This is another specific flow chart of step S400 of the data sensitivity determination method in the embodiment of the present application;

[0046] Figure 5 A specific flow chart of step S500 of the data sensitivity determination method in the embodiment of the present application;

[0047] Figure 6 This is another specific flow chart of the data sensitivity determination method in the embodiment of the present application;

[0048] Figure 7 A specific flow chart of step S600 of the data sensitivity determination method in the embodiment of the present application;

[0049] Figure 8 A specific application diagram of the data sensitivity determination method in an embodiment of the present application. DETAILED DESCRIPTION

[0050] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be understood as limiting the present application.

[0051] It should be noted that the logical order is shown in the flowchart, but in some cases, the steps shown or described can be performed in the order in different flowcharts. If it involves "several", it means more than one, if it involves "multiple", it means more than two, and the use of any and all examples or exemplary language ("for example", "such as", etc.) provided herein is intended only to better illustrate the embodiments of the present application, and unless otherwise required, it will not impose limitations on the scope of the present application. Greater than, less than, exceed, etc. are understood to exclude the number, and above, below, within, etc. are understood to include the number. If there is a description of the first and second, it is only for the purpose of distinguishing the technical features, and it cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0052] It should be noted that, unless otherwise specified, the singular forms of "a", "said" and "the" used in the embodiments are also intended to include plural forms, unless the context clearly indicates other meanings. In addition, unless otherwise defined, all technical and scientific terms used herein have the same meanings as those generally understood by those skilled in the art. The terms used in this specification are only for describing specific embodiments, not for limiting this application. The term "and / or" used herein includes any combination of one or more related listed items.

[0053] As the process of informatization gradually penetrates into daily life, digital storage of personal information, commercial secrets, and state secrets has become the mainstream, and the harm caused by various information leakage incidents has become more serious.

[0054] There are many main factors that cause data leakage at present. Taking the Internet industry, which accounts for the largest proportion, as an example, they include technical factors and non-technical factors. Among them, non-technical factors mainly include: internal employee leaks, poor internal information management, and human information leaks.

[0055] In order to strengthen the prevention of data leakage caused by non-technical reasons, the general solution is to implement sensitive data classification, information access permission control, information desensitization, encryption, watermarking, etc. Among them, taking sensitive data intelligent identification and classification technology as an example, only single data of sensitive information is analyzed, and a comprehensive analysis cannot be performed, resulting in low accuracy in sensitive information identification and failure to effectively prevent information leakage.

[0056] Specifically, the current identification of sensitive information in databases is based on analysis of single fields, but database access control technology rarely achieves access control at the field level. Although some database products can achieve field-level authorization, the maintenance and management workload is huge, and basically no organization can implement it. Therefore, only the sensitivity level and classification of the field are identified, and the accuracy of sensitive information identification is low, which cannot effectively prevent information leakage.

[0057] Based on this, the embodiments of the present application provide a data sensitivity determination method, an electronic device, and a computer-readable storage medium, which can comprehensively analyze sensitive information, improve the accuracy of identifying sensitive information, and effectively avoid the problem of information leakage.

[0058] In a first aspect, an embodiment of the present application provides a method for determining data sensitivity.

[0059] In some embodiments, reference Figure 1 , shows a flow chart of the data sensitivity determination method in the embodiment of the present application. It specifically includes the following steps:

[0060] S100, obtaining initial data to be determined;

[0061] S200, identifying the initial data to obtain multiple target fields of the initial data and field weights corresponding to the target fields;

[0062] S300, obtaining preset sensitive categories according to multiple target fields, and generating an initial matrix according to the multiple target fields and multiple field weights;

[0063] S400, calculating a scoring matrix of initial data according to a preset sensitive category, a preset weight database and an initial matrix;

[0064] S500: Determine the sensitivity of the initial data according to the scoring matrix.

[0065] In step S100, the initial data to be determined is obtained, wherein the initial data refers to data that may contain sensitive information and is collected through various data collection methods, such as data in a data table in a database or form data in a website interface, etc. In practical applications, multiple data records in each data table in a database can be collected in a random manner as the initial data to be determined, or each interface of the website can be randomly accessed, and DOM (Document Object Model) structure data in HTML (Hyper Text Markup Language) information can be obtained to collect each form data in the interface as the initial data to be determined.

[0066] In step S200, the initial data is identified to obtain multiple target fields in the initial data and field weights corresponding to the target fields, wherein the identification of the initial data is to identify whether there is sensitive information in the initial data, etc.; the target field refers to a sensitive field in the initial data that may be sensitive information; and the field weight refers to the possibility that the identified target field may be sensitive information, that is, the weight.

[0067] In some embodiments, when the initial data includes simple data, the simple data is identified based on the field features and regular expressions of multiple target fields to obtain multiple target fields and corresponding field weights of the simple data. Simple data refers to data with obvious information features of sensitive information in the initial data, such as simple data such as mobile phone numbers and ID card numbers, or data that is easy to identify such as field names in data tables; field features refer to preset fields similar to target sub-segments derived from target fields that need to be identified, such as field names for names, passwords, mobile phone numbers, etc. in data tables, and similar field features can be field features such as username, password, mobile, etc.; regular expressions are constructed in combination with field features to identify target fields. In addition, the target field identified by regular expressions and field features has a default default field weight of 1, that is, the target field is determined to be correct sensitive information. It should be noted that the number of field features in the embodiments of the present application is adjusted according to the number of target fields that need to be identified, and the present application does not make specific restrictions.

[0068] In some embodiments, when the initial data includes complex data, the complex data is identified based on a machine learning classification algorithm to obtain multiple target fields of the complex data and field weights corresponding to the target fields. Complex data refers to data in the initial data that has information features of sensitive information that is difficult to identify; a machine learning classification algorithm is a supervised learning method for modeling or predicting discrete random variables. The purpose of classification learning is to learn a classification function or classification model, also often referred to as a classifier, from a given manually annotated classification training sample data set. When new data arrives, predictions can be made based on this function to map the new data item to a class in a given category. The machine learning classification algorithm used in this application can be illustrated using the Bert algorithm as an example.

[0069] In practical applications, the algorithm model used for identification can be trained through a machine learning classification algorithm. The Chinese pre-trained library of Amazon AWS and the Bert machine learning classification algorithm are used for training. For example, the initial training data volume is about 100,000 of each type, and the verification set data volume is about 25,000. For example, model training is performed to generate a corresponding algorithm model. The algorithm model can identify the target fields and corresponding field weights in complex data. It should be noted that the embodiments of the present application do not describe the model training process in detail.

[0070] In some embodiments, the acquired initial data to be judged can be subjected to field processing, and identification marks that are beneficial to the machine learning classification algorithm can be added to determine the complex data in the initial data. Through this embodiment, the machine learning classification algorithm can quickly locate the complex data in the initial data and improve the recognition efficiency.

[0071] In steps S300 and S400, the preset sensitive categories are determined based on the identified multiple target fields, and an initial matrix is ​​generated based on the multiple target fields and the multiple field weights corresponding to the multiple target fields; then, a scoring matrix of the initial data is calculated based on the preset sensitive categories, the preset weight database and the initial matrix. The preset sensitive categories are obtained by the multiple target fields, and the initial matrix is ​​generated by the multiple target fields and the corresponding field weights; the preset weight database refers to the different preset weights that the multiple target fields have in different preset sensitive categories, and the size of the preset weight can be set according to actual needs, or by accumulating the weights of the multiple target fields previously obtained in the preset sensitive categories.

[0072] With respect to the preset sensitive categories, in the embodiment of the present application, the preset sensitive categories are categories summarized for multiple target fields, that is, when multiple target fields are identified in the initial data, the combination of the multiple target fields can determine the corresponding preset sensitive categories. For example, when the preset sensitive category is user identity information, the preset sensitive category includes but is not limited to any one or more target fields such as user name, ID number, passport, mobile phone number, address, driver's license, etc.; and the multiple target fields identified from the initial data are name, ID number, mobile phone number, address, and company name, then the combination of the multiple target fields can determine that the corresponding preset sensitive category is user identity information; it should be noted that the combination of multiple target fields can determine multiple different preset sensitive categories; and different preset sensitive categories have different sensitivities, that is, the sensitivity level of the data, which can be referred to. Figure 2 For example, there are extremely sensitive level, highly sensitive level, medium sensitive level, low sensitive level and non-sensitive level. Different sensitivity levels have different preset sensitive categories, and different preset sensitive categories involve different target fields.

[0073] It should be noted that Figure 2 The preset sensitive categories and corresponding sensitivities can be adjusted and set according to different business scenarios, and are not described or limited in detail in the embodiments of the present application.

[0074] The initial matrix is ​​explained. The initial matrix is ​​composed of multiple target fields and the field weights corresponding to the target fields. For example, when the multiple target fields to be identified are name, ID number, mobile phone number, address, and company name, and the corresponding field weights obtained after identification are 0.8, 0.0, 1.0, 0.87, and 0.0, the corresponding initial matrix is: [0.8 0.0 1.0 0.87 0.0]

[0076] The preset weight database is explained. The weight database contains several verification weight matrices, where the verification weight matrix is ​​a weight matrix of the score values, i.e., weight values, of the preset sensitive categories and corresponding target fields. The corresponding multiple verification weight matrices can be determined based on the obtained preset sensitive categories and target fields.

[0077] Correspondingly, in some embodiments, reference Figure 3 In step S400, the steps include:

[0078] S411, obtaining multiple verification weight matrices according to preset sensitive categories and a preset weight database;

[0079] S412, calculating a scoring matrix of the initial data based on the multiple verification weight matrices and the initial matrix.

[0080] In step S411, based on the identified multiple target fields, Figure 2 Determine the corresponding preset sensitive category, and then obtain the verification weight matrix corresponding to the preset sensitive category according to the preset weight database. In combination with actual application, multiple target fields can determine multiple corresponding preset sensitive categories. For example, when the multiple target fields identified are name, ID number, mobile phone number, address, and company name, the corresponding preset sensitive categories are user identity information, user private information, and legal person identity. The three preset sensitive categories combined with the weight database have corresponding verification weight matrices.

[0081] In step S412, the multiple verification weight matrices obtained and the initial matrices corresponding to the multiple target fields are combined for calculation to obtain a scoring matrix of the initial data. The specific calculation method can be calculated by the proximity distance algorithm WAKNN to obtain a scoring matrix of the multiple target fields. The scoring matrix is ​​a score for different preset sensitive categories, that is, the weight value of which preset sensitive category the multiple different target fields belong to.

[0082] For example, by taking the specific weight values ​​of the above initial matrix and multiple verification weight matrices as examples, calculations can be performed to obtain scoring matrices for different preset sensitive categories, and different preset sensitive categories have corresponding different scoring weights.

[0083] In some embodiments, reference Figure 4 In step S400, the steps include:

[0084] S421, acquiring a plurality of verification weight matrices and a preset weight matrix according to preset sensitive categories and a preset weight database;

[0085] S422, calculating a scoring matrix of the initial data according to a plurality of verification weight matrices, a preset weight matrix and an initial matrix.

[0086] In step S421, based on the identified multiple target fields, Figure 2 Determine the corresponding preset sensitive category, and then obtain the verification weight matrix and the corresponding preset weight matrix corresponding to the preset sensitive category from the preset weight database. The preset weight matrix refers to a weight matrix that sets the weight values ​​of multiple target fields in different preset sensitive categories according to actual conditions and needs. The preset weight matrix can adjust the calculation results of the verification weight matrix and the initial matrix to obtain a more accurate scoring matrix.

[0087] The target fields identified are name, ID number, mobile phone number, address, and company name. The corresponding preset sensitive categories are Figure 2 Taking user identity information, user private information, and legal person identity as an example, a verification weight matrix obtained for the three preset sensitive categories is: [0.8 0.8 0.3 0.3 0.2] [0.8 0.4 0.8 0.8 0.2] [0.2 0.6 0.6 0.7 0.8]

[0091] The meanings of the above three verification weight matrices are explained as follows: the weight values ​​of name, ID number, mobile phone number, address, and company name in user identity information are 0.8, 0.8, 0.3, 0.3, and 0.2 respectively; the weight values ​​of name, ID number, mobile phone number, address, and company name in user private information are 0.8, 0.4, 0.8, 0.8, and 0.2 respectively; the weight values ​​of name, ID number, mobile phone number, address, and company name in legal person identity identification are 0.2, 0.6, 0.6, 0.7, and 0.8 respectively.

[0092] Correspondingly, the initial matrix of multiple target fields is as follows: [0.8 0.0 1.0 0.87 0.0]

[0094] Correspondingly, the preset weight matrix is ​​as follows: [0.99 0.99 0.40 0.30 0.02] [0.90 0.18 0.85 0.82 0.02] [0.02 0.06 0.32 0.75 0.85]

[0098] The meaning of the above preset weight matrix is ​​explained as follows: the weight values ​​of name, ID number, mobile phone number, address, and company name in user identity information are 0.99, 0.99, 0.4, 0.3, and 0.02 respectively; the weight values ​​of name, ID number, mobile phone number, address, and company name in user private information are 0.9, 0.18, 0.85, 0.82, and 0.02 respectively; the weight values ​​of name, ID number, mobile phone number, address, and company name in legal person identity identification are 0.02, 0.06, 0.32, 0.75, and 0.85 respectively. The specific values ​​are obtained through analysis and experience accumulation, and their weight values ​​can also be adjusted.

[0099] It should be noted that in actual applications, the verification weight matrix includes multiple, and the specific number can be thousands or tens of thousands of verification weight matrices; and the number of selected preset weight matrices does not need to be too many, a small number will be enough to optimize the calculated scoring matrix. For example, in the embodiment of the present application, a preset weight matrix is ​​used as an example for illustration.

[0100] Based on the multiple verification weight matrices, initial matrices and weight matrices obtained above, the following scoring matrix is ​​obtained by calculating through the proximity distance algorithm WAKNN:

[0101] User identity information: [0.75]

[0102] User private information: [0.98]

[0103] Legal entity identity: [0.12]

[0104] To explain the above scoring matrix, that is, combining the obtained multiple target fields, for different preset sensitive categories, the weight value of user identity information, i.e. the similarity value, is 0.75, the weight value of user private information, i.e. the similarity value, is 0.98, and the weight value of legal person identity, i.e. the similarity value, is 0.12. Therefore, the user private information has the largest weight value, and the initial data is classified as user private information.

[0105] In step S500, the weight values ​​of the initial data in different preset sensitive categories are determined according to the obtained scoring matrix, the preset sensitive category to which the initial data belongs is determined by the weight value, and the corresponding sensitivity, i.e., the sensitivity level, is determined according to the different preset sensitive categories.

[0106] In some embodiments, reference Figure 5 In the embodiment of the present application, step S500 specifically further includes the steps of:

[0107] S510, determining the optimal sensitive category corresponding to the initial data according to the scoring matrix;

[0108] S520: Determine the sensitivity of the initial data according to the optimal sensitive category and a preset sensitive category database.

[0109] In step S510 and step S520, the specific weight values ​​of the initial data for different preset sensitive categories are determined based on the calculated scoring matrix, and the preset sensitive category with the largest weight value is selected as the optimal sensitive category, i.e., the highest priority preset sensitive category, and the sensitivity of the initial data is determined based on the sensitive positioning of the optimal sensitive category in the preset sensitive category database.

[0110] In some embodiments, reference Figure 6 , the embodiment of the present application specifically also includes the steps of:

[0111] S600: Adjust the weight of a preset weight matrix according to the scoring matrix.

[0112] In step S600, the preset weight matrix initially set is weight-adjusted according to the calculated scoring matrix to obtain an adjusted preset weight matrix, so that a more accurate scoring matrix can be obtained by calculating with the adjusted preset weight matrix. The weight adjustment refers to numerically adjusting each weight value in the preset weight matrix to ensure that the weight value is closer to the weight value of actual application. In actual application, the weight adjustment method can be numerically adjusted according to the scoring matrix and expert experience scoring method, or the adjusted weight matrix can be calculated by nonlinear least squares method.

[0113] In some embodiments, reference Figure 7 In the embodiment of the present application, step S600 specifically further includes:

[0114] S610, setting an expected matrix according to initial data;

[0115] S620, obtaining an error matrix based on the expectation matrix and the scoring matrix;

[0116] S630, calculating an increase vector matrix according to the error matrix and the preset weight matrix;

[0117] S640, calculating an adjusted preset weight matrix according to the increment matrix and the scoring matrix.

[0118] In step S610, according to the multiple target fields identified in the initial data and the corresponding preset sensitive categories, based on manual judgment or expert experience analysis, the expected matrix of the initial data for different preset sensitive categories, i.e., the expected scoring matrix, is set. For example, when the multiple target fields obtained after identifying the initial data are name, ID number, mobile phone number, address, company name, and the corresponding field weights are 0.8, 0.0, 1.0, 0.87, and 0.0, it means that the initial data center does not have identity-related information such as ID number and driver's license number, which does not meet the expected standard of user identity information, but it is determined that there are mobile phone numbers and addresses, which meet the standard of user private information. Based on this, the expected matrix, i.e., the expected matrix, is given by expert experience evaluation. For example, the expected matrix is ​​as follows:

[0119] User identity information: [0.25]

[0120] User private information: [0.98]

[0121] Legal entity identity: [0.12]

[0122] In step S620, an error matrix is ​​obtained according to the expected matrix and the scoring matrix calculated in step S400, wherein the error matrix refers to the error weight value between the expected matrix and the scoring matrix. For example, combining the scoring matrix and the expected matrix of the above example, the calculated error matrix is ​​as follows:

[0123] User identity information: [0.5]

[0124] User private information: [0.0]

[0125] Legal entity identity: [0.0]

[0126] In step S630, the calculated error matrix and the preset weight matrix are used to calculate to obtain a vector matrix, wherein the calculation method can be calculated by a nonlinear least square method, and the corresponding vector matrix is ​​obtained through multiple iterative calculations. For example, after combining the error matrix and the preset weight matrix of the above example for calculation, the corresponding vector matrix is ​​obtained as follows:

[0127] [-0.0002 0.001 -0.3 -0.03 0.001] [0.001 0.001 0.005 0.0 0.002]

[0129] [0.001 -0.001 0.003 0.001 0.001]

[0130] In step S640, the calculated increment matrix and the preset weight matrix are added to obtain an adjusted preset weight matrix. For example, after calculation in combination with the increment matrix and the preset weight matrix in the above example, the corresponding adjusted preset weight matrix is ​​obtained as follows: [0.9898 0.001 0.1 0.27 0.021] [0.981 0.181 0.855 0.82 0.022] [0.021 0.059 0.332 0.751 0.851]

[0134] In practical applications, the adjusted preset weight matrix can be used to calculate again in step S400 to obtain a new scoring matrix corresponding to the initial data. For example, after calculating in combination with the adjusted preset weight matrix of the above example, the corresponding new scoring matrix is ​​as follows:

[0135] User identity information: [0.281]

[0136] User private information: [0.985]

[0137] Legal entity identity: [0.123]

[0138] The accuracy of this scoring matrix is ​​higher than that of the originally calculated scoring matrix, which ensures the recognition accuracy of the sensitivity of the initial data.

[0139] In the possible application examples, refer to Figure 8, obtain the initial data waiting for determination in the data table or page form, identify the initial data, obtain the corresponding multiple target fields and field weights, and combine with the sensitive category database, determine the corresponding preset sensitive category and initial matrix according to the multiple target fields and the corresponding field weights, and combine with the weight database to perform matrix calculation to obtain the corresponding preset weight matrix and verification weight matrix, and after matrix calculation, obtain the scoring matrix corresponding to the initial data, and determine the sensitivity of the initial data according to the scoring matrix combined with the sensitive category database. It is also possible to adjust the weight of the preset weight matrix in the weight database according to the obtained scoring matrix and the corresponding sensitivity, obtain the adjusted preset weight matrix, and then perform matrix calculation to obtain a new scoring matrix to redetermine the sensitivity of the initial data.

[0140] In an embodiment of the present application, by acquiring initial data to be determined, the initial data is identified to obtain multiple target fields of the initial data and corresponding field weights, a preset sensitive category is acquired based on the multiple target fields, and an initial matrix is ​​generated based on the multiple target fields and the multiple field weights; based on the acquired preset sensitive categories, the preset weight database and the initial matrix, a scoring matrix of the initial data is calculated, and the sensitivity of the initial data is judged based on the scoring matrix, so that a comprehensive analysis of the sensitive information can be performed, the accuracy of identifying the sensitive information can be improved, and the problem of information leakage can be effectively avoided.

[0141] In a second aspect, an embodiment of the present application further provides an electronic device, comprising: at least one processor, and a memory communicatively connected to the at least one processor;

[0142] The processor is used to execute the data sensitivity determination method in the embodiment of the first aspect by calling the computer program stored in the memory.

[0143] The memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer executable programs, such as the data sensitivity determination method in the first embodiment of the present application. The processor implements the data sensitivity determination method in the first embodiment by running the non-transitory software programs and instructions stored in the memory.

[0144] The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system and application programs required for at least one function; the data storage area may store and execute the data sensitivity determination method in the first aspect embodiment. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0145] The non-transient software programs and instructions required to implement the data sensitivity determination method in the above-mentioned first aspect embodiment are stored in the memory, and when executed by one or more processors, the data sensitivity determination method in the above-mentioned first aspect embodiment is executed.

[0146] In a third aspect, an embodiment of the present application further provides a computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are used to: execute the data sensitivity determination method in the embodiment of the first aspect;

[0147] In some embodiments, the computer-readable storage medium stores computer-executable instructions, which are executed by one or more control processors, for example, by a processor in the electronic device of the second aspect embodiment, so that the one or more processors can execute the data sensitivity determination method in the first aspect embodiment.

[0148] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0149] It will be appreciated by those skilled in the art that all or some of the steps and systems in the disclosed method above may be implemented as software, firmware, hardware and appropriate combinations thereof. Some physical components or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor or a microprocessor, or may be implemented as hardware, or may be implemented as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or a non-transitory medium) and a communication medium (or a temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that may be used to store desired information and may be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0150] In the description of this specification, the description with reference to the terms "some embodiments", "examples", "specific examples", or "some examples" means that the specific features or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example.

Claims

1. A method for determining data sensitivity, characterized in that: include: Obtaining initial data to be determined; Identify the initial data to obtain multiple target fields of the initial data and field weights corresponding to the target fields; Acquire preset sensitive categories according to the multiple target fields, and generate an initial matrix according to the multiple target fields and multiple field weights; A scoring matrix of the initial data is calculated according to the preset sensitive category, the preset weight database and the initial matrix; wherein the weight database includes a plurality of verification weight matrices, and the verification weight matrix represents the score value of the sensitive category and the corresponding target field; Determining the sensitivity of the initial data according to the scoring matrix; The step of calculating the scoring matrix of the initial data according to the preset sensitive category, the preset weight database and the initial matrix includes: Acquire multiple verification weight matrices according to the preset sensitive categories and the preset weight database; Calculating the scoring matrix of the initial data by using a proximity distance algorithm according to the multiple verification weight matrices and the initial matrix; The determining the sensitivity of the initial data according to the scoring matrix includes: Determining the optimal sensitive category corresponding to the initial data according to the scoring matrix; The sensitivity of the initial data is determined according to the optimal sensitive category and a preset sensitive category database.

2. The data sensitivity determination method according to claim 1, characterized in that: The initial data includes simple data; The identifying the initial data to obtain a plurality of target fields of the initial data and field weights corresponding to the target fields includes: The simple data is identified based on field features of multiple target fields and regular expressions to obtain multiple target fields of the simple data and field weights corresponding to the target fields.

3. The data sensitivity determination method according to claim 2, characterized in that: The initial data includes complex data; The identifying the initial data to obtain a plurality of target fields of the initial data and field weights corresponding to the target fields includes: The complex data is identified based on a machine learning classification algorithm to obtain multiple target fields of the complex data and field weights corresponding to the target fields.

4. The data sensitivity determination method according to claim 2 or 3, characterized in that: The weight database includes a plurality of the verification weight matrices and a preset weight matrix, wherein the preset weight matrix represents weight values ​​of the plurality of target fields in different preset sensitive categories, and the preset weight matrix can optimize the calculation results of the verification weight matrix and the initial matrix; Correspondingly, the scoring matrix of the initial data is obtained by calculating according to the preset sensitive category, the preset weight database and the initial matrix, including: Acquire a plurality of verification weight matrices and a preset weight matrix according to the preset sensitive category and the preset weight database; A scoring matrix for the initial data is obtained by calculating the multiple verification weight matrices, the preset weight matrix and the initial matrix using a proximity distance algorithm.

5. The data sensitivity determination method according to claim 4, characterized in that: Also includes: The preset weight matrix is ​​weight-adjusted according to the scoring matrix.

6. The data sensitivity determination method according to claim 5, characterized in that: The preset weight matrix is ​​weight-adjusted according to the scoring matrix, further comprising: Setting an expectation matrix according to the initial data; Acquire an error matrix based on the expected matrix and the scoring matrix; Calculate a vector matrix based on the error matrix and the preset weight matrix; An adjusted preset weight matrix is ​​calculated according to the multiplication factor matrix and the preset weight matrix.

7. An electronic device, characterized in that include: at least one processor, and a memory communicatively connected to the at least one processor; wherein, The memory stores instructions, and the instructions are executed by the at least one processor so that the at least one processor implements the data sensitivity determination method according to any one of claims 1 to 6 when executing the instructions.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the data sensitivity determination method according to any one of claims 1 to 6.

Images