Data security management and control system based on data classification and grading method

Abstract

The invention relates to the technical field of data classification and grading, and discloses a data security management and control system based on a data classification and grading method, which comprises the steps of classifying to-be-processed data, then respectively processing according to different data categories, and when the data categories are personal data, storing the personal data. Determining a corresponding individual grading rule based on the first economic value feature and the legal association feature, when the data category is enterprise data, determining a corresponding enterprise grading rule based on the second economic value feature and the life cycle feature, and when the data category is public data, determining a corresponding enterprise grading rule. And determining a corresponding public grading rule according to the leakage characteristic and the third economic value characteristic, and adjusting the grading rule in real time according to the economic value characteristic, the law association characteristic, the leakage characteristic and other dynamic factors of the actual data, so that the data management is more flexible, and different application scenes and market changes can be quickly adapted.

Description

Technical Field

[0001] The present invention relates to the technical field of data classification and grading, and more specifically, to a data security management and control system based on a data classification and grading method. Background Art

[0002] Data classification and grading is essentially the process of labeling and symbolizing data based on the intrinsic attributes, characteristics and human purposes of the data. Specifically, data classification is the process of classifying similar or common data according to the attributes and characteristics of the data, and establishing a hierarchical classification system in accordance with certain principles and methods. Data grading is based on data classification and assigns different protection levels or processing priorities to data according to its sensitivity, importance, value and other characteristics to ensure that the data can be properly managed and protected. Although relevant content about data classification and grading has appeared in the prior art, there are still certain problems.

[0003] For example, the patent with publication number CN115687725A provides a method and device for data classification and grading. The patent determines the level of data based on the classification and grading rules and the category of data. The patent with publication number CN116541573A provides a data management method based on hierarchical classification. The patent calculates the privacy hierarchical classification attributes of the data and forms a set of digital coding rules for data classification based on pre-configured hierarchical classification standards.

[0004] Although the above-mentioned patents all provide relevant content on data classification and grading, the classification and grading rules of the above-mentioned patents are all pre-set and fail to fully consider the dynamic characteristics of data and changes in the market environment. This fixed rule approach may lead to insufficient flexibility in data management when faced with complex and changeable actual application scenarios, and it is difficult to accurately respond to the specific needs of different types of data.

[0005] In view of this, the present invention proposes a data security management and control system based on a data classification and grading method to solve the above problems. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art, the present invention provides a data security management and control system based on a data classification and grading method.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] In a first aspect, a data security management and control system based on a data classification and grading method comprises:

[0009] Classification module: used to obtain the data to be processed, input the data to be processed into the pre-built data classification model, and obtain the data category, which includes personal data, enterprise data and public data;

[0010] The first classification module is used for obtaining the first economic value feature and the legal association feature according to the personal data when the data category is personal data, and determining the corresponding personal classification rules based on the first economic value feature and the legal association feature, wherein the personal classification rules at least include responsibility rules and property rules;

[0011] The second classification module is used for obtaining the second economic value characteristics and life cycle characteristics according to the enterprise data when the data category is enterprise data, and determining the corresponding enterprise classification rules based on the second economic value characteristics and life cycle characteristics. The enterprise classification rules at least include life cycle rules and property rules;

[0012] The third classification module: when the data category is public data, it obtains the leakage characteristics and the third economic value characteristics according to the public data, and determines the corresponding public classification rules according to the leakage characteristics and the third economic value characteristics. The public classification rules at least include leakage rules and property rules.

[0013] Furthermore, the method for constructing the data classification model includes:

[0014] Get r groups of data, where r is a positive integer greater than 1, and the data include historical data to be processed and historical data categories. Take the historical data to be processed and the historical data categories as sample sets, divide the sample sets into training sets and test sets, build a classifier, take the historical data to be processed in the training set as input data, take the historical data categories in the training set as output data, train the classifier to obtain an initial classifier, test the initial classifier with the test set, and output a classifier that meets the preset accuracy as a data classification model.

[0015] Furthermore, the method for obtaining the first economic value feature based on personal data includes:

[0016] The personal data is clustered according to the economic type to generate a first clustering result, and the first value characteristic is determined based on the first clustering result and the corresponding first influencing factor. The personal data is clustered according to the time period to generate a second clustering result, and the second value characteristic is determined based on the second clustering result and the corresponding second influencing factor. The personal data is clustered according to the data usage scope to generate a third clustering result, and the third value characteristic is determined based on the third clustering result and the corresponding third influencing factor. The first economic value characteristic is calculated based on the first value characteristic, the second value characteristic and the third value characteristic.

[0017] Furthermore, the method for calculating the first economic value feature based on the first value feature, the second value feature and the third value feature includes:

[0018]

[0019] In the formula, EVF is the first economic value feature, FVF is the first value feature, SVF is the second value feature, TVF is the third value feature, cot -1 (·) is the inverse cotangent function, cos -1 [·] is the arccosine function, sin -1 [·] is the inverse sine function, exp(·) is the exponential function with e as the base, log 10 (·) is a logarithmic function with base 10, and K is a constant greater than 0.

[0020] Furthermore, methods for obtaining legal association characteristics based on personal data include:

[0021] According to the pre-built keyword extraction model, N target keywords are extracted from personal data. The target knowledge graph is constructed with the N target keywords as nodes and the relationships between the N target keywords as edges. The complexity of the target knowledge graph is calculated and the complexity is used as the legal association feature.

[0022] Furthermore, the method for calculating the complexity of the target knowledge graph includes:

[0023]

[0024] Where Cplx is the complexity, N is the number of nodes in the target knowledge graph, and E is the number of edges in the target knowledge graph.

[0025] Furthermore, the method for determining corresponding individual classification rules based on the first economic value feature and the legal association feature includes:

[0026] The first economic value feature and the legal association feature are input into a pre-constructed first grading determination model to obtain a personal grading rule.

[0027] The method for constructing the first hierarchical determination model includes:

[0028] A sample data set is obtained, wherein the sample data set includes a historical first economic value feature, a historical legal association feature, and a historical personal grading rule; the sample data set is divided into a sample training set and a sample test set; a regression network is constructed; the historical first economic value feature and the historical legal association feature in the sample training set are used as input data of the regression network; the historical personal grading rules in the sample training set are used as output data of the regression network; the regression network is trained to obtain an initial regression network for predicting real-time personal grading rules; the initial regression network is tested using the sample test set; and the initial regression network that satisfies a preset error value is output as a first grading determination model.

[0029] Furthermore, the method of obtaining life cycle characteristics based on enterprise data includes:

[0030] Preset the life cycle stage, determine the number of elements in the enterprise data that are in the life cycle stage, and use the number of elements in the life cycle stage as the life cycle feature. The life cycle stage at least includes the data storage stage, data usage stage, data transmission stage and data release stage.

[0031] Furthermore, the method for obtaining leakage characteristics based on public data includes:

[0032] Obtain relevant sensitive data of public data, calculate the leakage risk coefficient based on the relevant sensitive data, and use the leakage risk coefficient as the leakage feature. The relevant sensitive data at least includes the number of access exceptions, the number of transmission path nodes, the number of target entities, and the historical leakage recovery cost.

[0033] In a second aspect, a data security management and control method based on a data classification and grading method is implemented based on the above-mentioned data security management and control system based on a data classification and grading method, including:

[0034] S10: Obtain the data to be processed, input the data to be processed into a pre-built data classification model, and obtain data categories, where the data categories include personal data, enterprise data, and public data;

[0035] S20: When the data category is personal data, obtaining a first economic value feature and a legal association feature according to the personal data, and determining a corresponding personal classification rule based on the first economic value feature and the legal association feature, wherein the personal classification rule at least includes a responsibility rule and a property rule;

[0036] S30: When the data category is enterprise data, a second economic value feature and a life cycle feature are obtained according to the enterprise data, and a corresponding enterprise classification rule is determined based on the second economic value feature and the life cycle feature. The enterprise classification rule at least includes a life cycle rule and a property rule.

[0037] S40: When the data category is public data, the leakage characteristics and the third economic value characteristics are obtained according to the public data, and the corresponding public classification rules are determined according to the leakage characteristics and the third economic value characteristics. The public classification rules at least include leakage rules and property rules.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] In the present invention, the data to be processed is first classified, and then processed according to different data categories. When the data category is personal data, the corresponding personal classification rules are determined based on the first economic value feature and the legal association feature. When the data category is enterprise data, the corresponding enterprise classification rules are determined based on the second economic value feature and the life cycle feature. When the data category is public data, the corresponding public classification rules are determined according to the leakage feature and the third economic value feature. In the present invention, the classification rules are adjusted in real time according to dynamic factors such as the economic value characteristics, legal association characteristics, and leakage characteristics of the actual data, so that data management is more flexible and can quickly adapt to different application scenarios and market changes. Since different data have different characteristics and risks, this embodiment can provide personalized data classification rules and formulate special management strategies for different data categories (personal data, enterprise data, public data). BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a structural schematic diagram of a data security management and control system based on a data classification and grading method in the present invention;

[0041] Figure 2 This is a flow chart of a data security management and control method based on a data classification and grading method in the present invention;

[0042] Figure 3 A flowchart of a method for obtaining a first economic value feature based on personal data in the present invention;

[0043] Figure 4 A schematic diagram of a computer-readable storage medium in the present invention. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] Example 1

[0046] See also Figure 1As shown, this embodiment discloses a data security management and control system based on a data classification and grading method, including:

[0047] Classification module: used to obtain the data to be processed, input the data to be processed into the pre-built data classification model, and obtain the data category, which includes personal data, enterprise data and public data;

[0048] In this embodiment, the data to be processed may include only one type of data, such as personal data, or may include personal data and corporate data, etc. In the process of classifying the data to be processed, the data to be processed is also divided. It can be understood that before the data to be processed is classified, the data to be processed may also be preprocessed. The preprocessing includes missing value processing, duplicate value deletion, outlier processing, and consistency check, etc. This embodiment does not go into too much detail on this.

[0049] Personal data refers to any information related to a specific individual, whether the information is directly or indirectly associated with the individual's identity. This data can be used to identify a specific individual and involves the individual's privacy and security. Enterprise data refers to data related to enterprise operations, management, and business activities. This data involves the core assets and trade secrets of the enterprise. Public data refers to data that is usually generated and maintained by the government or public institutions. This data is not specific to individuals or enterprises, but serves the public and society broadly.

[0050] The methods for building data classification models include:

[0051] Obtain r groups of data, where r is a positive integer greater than 1, and the data includes historical data to be processed and historical data categories. The historical data to be processed and the historical data categories are used as sample sets, and the sample sets are divided into training sets and test sets. A classifier is constructed, and the historical data to be processed in the training set is used as input data, and the historical data categories in the training set are used as output data. The classifier is trained to obtain an initial classifier, and the initial classifier is tested using the test set. A classifier that meets a preset accuracy is output as a data classification model, and the classifier is preferably a naive Bayes model or a support vector machine model.

[0052] It should be added that, in this embodiment, the data to be processed is classified according to the data source. The purpose of this is to classify the data as much as possible according to market rules, so that the classification and subsequent processing of the data can be closer to the actual application scenario.

[0053] The first classification module is used for obtaining the first economic value feature and the legal association feature according to the personal data when the data category is personal data, and determining the corresponding personal classification rules based on the first economic value feature and the legal association feature, wherein the personal classification rules at least include responsibility rules and property rules;

[0054] See also Figure 3 As shown, in this embodiment, the method for obtaining the first economic value feature according to personal data includes:

[0055] The personal data is clustered according to the economic type to generate a first clustering result, and the first value characteristic is determined based on the first clustering result and the corresponding first influencing factor. The personal data is clustered according to the time period to generate a second clustering result, and the second value characteristic is determined based on the second clustering result and the corresponding second influencing factor. The personal data is clustered according to the data usage scope to generate a third clustering result, and the third value characteristic is determined based on the third clustering result and the corresponding third influencing factor. The first economic value characteristic is calculated based on the first value characteristic, the second value characteristic and the third value characteristic.

[0056] It should be noted that the economic type includes at least financial type, social type and consumption type. Exemplarily, personal data related to the financial type may be income records, expenditure records, investment records, bank account information, etc., and personal data related to the social type may be social media activities, contact networks, social interaction records, etc. It can be understood that the clustering result in this embodiment can be the data volume. For example, after the personal data is vectorized, the data volume can be the length of the vector. The above-mentioned clustering can be achieved through existing K-means clustering or hierarchical clustering, which is not elaborated in this embodiment.

[0057] Taking the first clustering result as an example, the method for determining the first value feature based on the first clustering result and the corresponding first influencing factor includes: multiplying the data volume of the first clustering result by its corresponding first influencing factor to obtain the first value feature. It can be understood that the first influencing factor corresponding to the financial type is greater than the first influencing factor corresponding to the consumption type, and the first influencing factor corresponding to the consumption type is greater than the first influencing factor corresponding to the social type.

[0058] The above-mentioned time interval refers to the time interval when the personal data is generated. The personal data is clustered through the preset time intervals to generate the second clustering result. It can be understood that the closer the time interval is to the current time point, the greater the value of the corresponding second impact factor. This is because the timeliness of personal data is very critical in economic value assessment. The closer the data is to the current time, the more practical economic value and reference significance it usually has.

[0059] The above-mentioned data usage scope refers to the range of the product of the number of data usage times and the number of data scenarios. The number of data usage times refers to the number of times the data is actually used or queried, and the number of data scenarios refers to the number of data in different applications or scenarios. The data usage scope characterizes the overall usage frequency and application breadth of the data. Data with high usage frequency and wide application show the actual value and importance of the data, so these data often have higher economic value. Therefore, in this embodiment, the larger the data usage scope, the larger the value of the corresponding third impact factor.

[0060] The method for calculating the first economic value feature based on the first value feature, the second value feature and the third value feature includes:

[0061]

[0062] In the formula, EVF is the first economic value feature, FVF is the first value feature, SVF is the second value feature, TVF is the third value feature, cot -1 (·) is the inverse cotangent function, cos -1 [·] is the arccosine function, sin -1 [·] is the inverse sine function, exp(·) is the exponential function with e as the base, log 10 (·) is a logarithmic function with base 10, and K is a constant greater than 0.

[0063] In this embodiment, taking the first value characteristic as an example, the larger the first value characteristic is, the larger the corresponding first economic value characteristic is. Therefore, the first economic value characteristic is positively correlated with the economic value of personal data. Therefore, the larger the first economic value characteristic is, the more it is necessary to classify the personal data according to property rules.

[0064] Methods for obtaining legally relevant characteristics from personal data include:

[0065] According to the pre-built keyword extraction model, N target keywords are extracted from personal data. The target knowledge graph is constructed with the N target keywords as nodes and the relationships between the N target keywords as edges. The complexity of the target knowledge graph is calculated and the complexity is used as the legal association feature.

[0066] It should be noted that the keyword extraction model can be trained through existing models such as Word2Vec and FastText. The training process is to input the preset keyword library into the above-mentioned existing model, so as to obtain the keyword extraction model in this embodiment. It is worth noting that the above-mentioned keyword library refers to a keyword library related to law.

[0067] It should be added that the keyword library includes ID card number, bank account information, credit card number, medical records, diagnostic information, fingerprints, facial recognition data, etc. It is understandable that once the above keywords are leaked, the corresponding responsibility can be pursued according to the law. Therefore, the above keywords have a strong correlation with the law. The relationship between the above target keywords at least includes dependency, logical relationship and hierarchical relationship. For example, the dependency relationship is that there is a dependency relationship between "credit card number" and "bank account information", because credit card information usually relies on bank account information to complete transactions. For example, the logical relationship is that "ID card number" and "facial recognition data" are used together for identity authentication, and there is a logical dependency relationship between them.

[0068] Methods for calculating the complexity of the target knowledge graph include:

[0069]

[0070] Where Cplx is the complexity, N is the number of nodes in the target knowledge graph, and E is the number of edges in the target knowledge graph.

[0071] In this embodiment, complexity is defined as the relationship between nodes and edges in the target knowledge graph. In the knowledge graph, each node represents a target keyword (such as ID number, bank account information, etc.), and each edge represents the relationship between these keywords (such as dependency, logical relationship, hierarchical relationship). When the complexity of the target knowledge graph is high, it means that there are more associations and interactions between the key legal information involved in the personal data. This high degree of correlation indicates that multiple important data points are interdependent or work together, which usually means that once these data are leaked, they may lead to more complex legal issues. For example, when ID number and facial recognition data are leaked at the same time, the identity security of the victim will be more threatened, and the legal responsibilities involved will be more complex and serious. As the complexity of the target knowledge graph increases, the sensitivity and importance of personal data are also increasing. Therefore, it can be understood that the greater the complexity, the stronger the legal nature of the personal data, and the more necessary it is to classify the personal data according to the responsibility rules. The responsibility here refers to legal responsibility.

[0072] The method for determining the corresponding individual classification rules based on the first economic value feature and the legal association feature includes:

[0073] The first economic value feature and the legal association feature are input into a pre-constructed first grading determination model to obtain a personal grading rule.

[0074] The method for constructing the first hierarchical determination model includes:

[0075] A sample data set is obtained, wherein the sample data set includes historical first economic value features, historical legal association features, and historical personal grading rules; the sample data set is divided into a sample training set and a sample test set; a regression network is constructed; the historical first economic value features and historical legal association features in the sample training set are used as input data of the regression network; the historical personal grading rules in the sample training set are used as output data of the regression network; the regression network is trained to obtain an initial regression network for predicting real-time personal grading rules; the initial regression network is tested using the sample test set; and the initial regression network that satisfies a preset error value is output as a first grading determination model; the initial regression network is preferably a deep neural network model.

[0076] It should be noted that the above-mentioned grading determination model is generated by supervised learning, that is, the importance of the first economic value feature and the legal association feature is compared based on expert experience, so as to select the corresponding personal grading rules. For example, when the first economic value feature is more important, the property rule in the personal grading rules is selected. The above-mentioned method for obtaining importance can be to first preset the corresponding feature standard value. For example, the ratio of the first economic value feature to the economic feature standard value is used to characterize the importance.

[0077] It should be added that the responsibility rules mentioned above refer to a type of hierarchical rules, which clarify the legal compliance and accountability mechanism in the data processing process through grading to protect the rights and interests of data subjects, and the property rules focus on economic value, protect the value of data as an asset through grading, and ensure that data is not abused or infringed during use and sharing.

[0078] The second classification module is used for obtaining the second economic value characteristics and life cycle characteristics according to the enterprise data when the data category is enterprise data, and determining the corresponding enterprise classification rules based on the second economic value characteristics and life cycle characteristics. The enterprise classification rules at least include life cycle rules and property rules;

[0079] It can be understood that the method for obtaining the second economic value feature based on enterprise data can be the same as the method for obtaining the first economic value feature in the above steps, except that there are some differences in the data. For example, the economic type includes at least financial type, social type and consumption type. For example, personal data related to the financial type can be income records, expenditure records, investment records, bank account information, etc., and personal data related to the social type can be social media activities, contact networks, social interaction records, etc. Then, relative to the enterprise, the enterprise data related to the financial type can be a balance sheet, an income statement, a cash flow statement, a tax return, a tax compliance record, etc., and the enterprise data related to the social type can be advertising records, social media marketing activities, marketing strategies, internal corporate emails, meeting minutes, etc. The specific calculation method can be the same, so this embodiment does not go into too much detail.

[0080] Methods for obtaining life cycle characteristics based on enterprise data include:

[0081] Preset the life cycle stage, determine the number of elements in the enterprise data that are in the life cycle stage, and use the number of elements in the life cycle stage as the life cycle feature. The life cycle stage at least includes the data storage stage, data usage stage, data transmission stage and data release stage.

[0082] It should be noted that the life cycle mentioned above refers to the life cycle of enterprise data. Therefore, under normal circumstances, enterprise data will face different risks and management needs at each stage of its life cycle. Therefore, in this embodiment, the larger the life cycle characteristics corresponding to the enterprise data, the more it indicates that the enterprise data needs to be classified according to the life cycle rules.

[0083] In the above description, each element in the enterprise data refers to each piece of data in the enterprise data. Then, the method for determining the number of each element in the enterprise data in the life cycle stage includes:

[0084] Traverse each element in the enterprise data, determine whether it is in the data storage stage according to the storage time of each element, determine whether it is in the data storage stage according to the access time of each element, determine whether it is in the data transmission stage according to the transmission path of each element, and determine whether it is in the data release stage according to the release time of each element.

[0085] It should be added that when an element is in any sub-stage, the element is in the life cycle stage. For example, the data storage stage is a sub-stage of the life cycle stage. When the storage time corresponding to the element is detected, it is judged that the element is in the data storage stage. Accordingly, the element is in the life cycle stage.

[0086] In this embodiment, when the life cycle characteristics corresponding to the enterprise data are larger, it means that more elements are active in the various stages of the enterprise data's life cycle (such as data storage, data usage, data transmission, and data release). This characteristic reflects the activity level and complexity of the enterprise data as well as the risks and management needs that may be faced at different stages. Therefore, when the life cycle characteristics corresponding to the enterprise data are larger, it indicates that the enterprise data needs to be classified according to the life cycle rules.

[0087] The method for determining the corresponding enterprise classification rules based on the second economic value characteristics and life cycle characteristics includes:

[0088] The second economic value characteristics and the life cycle characteristics are input into the pre-built second classification determination model to obtain the enterprise classification rules.

[0089] Similarly, the second grading determination model in this embodiment is constructed in the same way as the first grading determination model in the above steps, except that the input data is changed to the second economic value characteristics and life cycle characteristics, and the output data is changed to the enterprise grading rules.

[0090] It should be added that life cycle rules refer to the rules used in enterprise data management to formulate corresponding management and protection strategies based on the different stages of data in its life cycle.

[0091] The third classification module is used for obtaining the leakage characteristics and the third economic value characteristics according to the public data when the data category is public data, and determining the corresponding public classification rules according to the leakage characteristics and the third economic value characteristics, wherein the public classification rules at least include the leakage rules and the property rules;

[0092] It can be understood that the method for obtaining the third economic value feature based on public data can be the same as the method for obtaining the first economic value feature in the above content, except that there are some differences in the data. For example, the economic type includes at least financial type, social type and consumption type. For example, personal data related to the financial type can be income records, expenditure records, investment records, bank account information, etc., and personal data related to the social type can be social media activities, contact networks, social interaction records, etc. Then, relative to public institutions, public data related to the financial type can be the annual fiscal budget of the government or public institution, expenditure details of various departments, project grant records, tax revenue data of public institutions, etc., and public data related to the social type can be public opinions and consultation records collected by public institutions through various channels, content published by public institutions on social media platforms, interaction records, etc. The specific calculation method can be the same, so this embodiment does not go into too much detail.

[0093] It should be added that leakage rules refer to the grading rules used to deal with data leakage risks in public data management. These rules classify data and formulate corresponding management and protection measures based on factors such as the possibility of data leakage, scope of impact, and recovery costs.

[0094] Methods for obtaining leakage characteristics based on public data include:

[0095] Obtain relevant sensitive data of public data, calculate the leakage risk coefficient based on the relevant sensitive data, and use the leakage risk coefficient as the leakage feature. The relevant sensitive data at least includes the number of access anomalies, the number of transmission path nodes, the number of target entities, and the historical leakage recovery cost;

[0096] It can be understood that the number of abnormal access times refers to the number of times public data is accessed at abnormal times or abnormal geographical locations, the number of transmission path nodes refers to the number of nodes involved in public data transmission, the number of target entities refers to the number of individuals or institutions affected by public data, and the leakage recovery cost refers to the cost of recovery after public data leakage. Taking the number of abnormal access times and the number of target entities as examples, frequent abnormal access indicates that public data may be at risk, and the larger the number of target entities, the more people or organizations will be affected after the data leakage, and the more serious the consequences will be.

[0097] Methods for calculating leakage risk factors based on relevant sensitive data include:

[0098]

[0099] Where LRC is the leakage risk coefficient, NAS is the number of access anomalies, NTP is the number of transmission path nodes, NTE is the number of target entities, cosh(·) is the hyperbolic cosine function, sinh[·] is the hyperbolic sine function, and 1! is the factorial of 1.

[0100] In this embodiment, taking the number of access anomalies as an example, the greater the number of access anomalies, the greater the corresponding leakage risk coefficient. Since the leakage risk coefficient is used as a leakage feature, the leakage feature is positively correlated with the leakage risk level of public data. Therefore, the greater the leakage feature, the more it is necessary to classify the public data according to the leakage rules.

[0101] The method for determining corresponding public classification rules according to the leakage characteristics and the third economic value characteristics includes:

[0102] The leakage characteristics and the third economic value characteristics are input into a pre-built third grading determination model to obtain a common grading rule.

[0103] Similarly, the third grading determination model in this embodiment is constructed in the same way as the first grading determination model in the above steps, except that the input data is changed to leakage characteristics and the third economic value characteristics, and the output data is changed to public grading rules. This embodiment does not go into details about this.

[0104] In this embodiment, the data to be processed is first classified and then processed according to different data categories. When the data category is personal data, the corresponding personal classification rules are determined based on the first economic value characteristics and the legal association characteristics. When the data category is enterprise data, the corresponding enterprise classification rules are determined based on the second economic value characteristics and the life cycle characteristics. When the data category is public data, the corresponding public classification rules are determined according to the leakage characteristics and the third economic value characteristics. In this embodiment, the classification rules are adjusted in real time according to dynamic factors such as the economic value characteristics, legal association characteristics, and leakage characteristics of the actual data, so that data management is more flexible and can quickly adapt to different application scenarios and market changes. Since different data have different characteristics and risks, this embodiment can provide personalized data classification rules and formulate special management strategies for different data categories (personal data, enterprise data, public data).

[0105] Example 2

[0106] See also Figure 2 As shown, this embodiment provides a data security management method based on the data classification and grading method on the basis of embodiment 1, including:

[0107] S10: Obtain the data to be processed, input the data to be processed into a pre-built data classification model, and obtain data categories, where the data categories include personal data, enterprise data, and public data;

[0108] In this embodiment, the data to be processed may include only one type of data, such as personal data, or may include personal data and corporate data, etc. In the process of classifying the data to be processed, this embodiment also divides the data to be processed. It can be understood that before classifying the data to be processed, the data to be processed may also be preprocessed. The preprocessing includes missing value processing, duplicate value deletion, outlier processing, and consistency check, etc. This embodiment does not go into too much detail on this.

[0109] The methods for building data classification models include:

[0110] Obtain r groups of data, where r is a positive integer greater than 1, and the data includes historical data to be processed and historical data categories. The historical data to be processed and the historical data categories are used as sample sets, and the sample sets are divided into training sets and test sets. A classifier is constructed, and the historical data to be processed in the training set is used as input data, and the historical data categories in the training set are used as output data. The classifier is trained to obtain an initial classifier, and the initial classifier is tested using the test set. A classifier that meets a preset accuracy is output as a data classification model, and the classifier is preferably a naive Bayes model or a support vector machine model.

[0111] S20: When the data category is personal data, obtaining a first economic value feature and a legal association feature according to the personal data, and determining a corresponding personal classification rule based on the first economic value feature and the legal association feature, wherein the personal classification rule at least includes a responsibility rule and a property rule;

[0112] In this embodiment, the method for obtaining the first economic value feature based on personal data includes:

[0113] The personal data is clustered according to the economic type to generate a first clustering result, and the first value characteristic is determined based on the first clustering result and the corresponding first influencing factor. The personal data is clustered according to the time period to generate a second clustering result, and the second value characteristic is determined based on the second clustering result and the corresponding second influencing factor. The personal data is clustered according to the data usage scope to generate a third clustering result, and the third value characteristic is determined based on the third clustering result and the corresponding third influencing factor. The first economic value characteristic is calculated based on the first value characteristic, the second value characteristic and the third value characteristic.

[0114] The method for calculating the first economic value feature based on the first value feature, the second value feature and the third value feature includes:

[0115]

[0116] In the formula, EVF is the first economic value feature, FVF is the first value feature, SVF is the second value feature, TVF is the third value feature, cot -1 (·) is the inverse cotangent function, cos -1 [·] is the arccosine function, sin -1 [·] is the inverse sine function, exp(·) is the exponential function with e as the base, log 10 (·) is a logarithmic function with base 10, and K is a constant greater than 0.

[0117] In this embodiment, taking the first value characteristic as an example, the larger the first value characteristic is, the larger the corresponding first economic value characteristic is. Therefore, the first economic value characteristic is positively correlated with the economic value of personal data. Therefore, the larger the first economic value characteristic is, the more it is necessary to classify the personal data according to property rules.

[0118] Methods for obtaining legally relevant characteristics based on personal data include:

[0119] According to the pre-built keyword extraction model, N target keywords are extracted from personal data. The target knowledge graph is constructed with the N target keywords as nodes and the relationships between the N target keywords as edges. The complexity of the target knowledge graph is calculated and the complexity is used as the legal association feature.

[0120] It should be noted that the keyword extraction model can be trained through existing models such as Word2Vec and FastText. The training process is to input the preset keyword library into the above-mentioned existing model, so as to obtain the keyword extraction model in this embodiment. It is worth noting that the above-mentioned keyword library refers to a keyword library related to law.

[0121] Methods for calculating the complexity of the target knowledge graph include:

[0122]

[0123] Where Cplx is the complexity, N is the number of nodes in the target knowledge graph, and E is the number of edges in the target knowledge graph.

[0124] The method for determining the corresponding individual classification rules based on the first economic value feature and the legal association feature includes:

[0125] The first economic value feature and the legal association feature are input into a pre-constructed first grading determination model to obtain a personal grading rule.

[0126] The method for constructing the first hierarchical determination model includes:

[0127] A sample data set is obtained, wherein the sample data set includes historical first economic value features, historical legal association features, and historical personal grading rules; the sample data set is divided into a sample training set and a sample test set; a regression network is constructed; the historical first economic value features and historical legal association features in the sample training set are used as input data of the regression network; the historical personal grading rules in the sample training set are used as output data of the regression network; the regression network is trained to obtain an initial regression network for predicting real-time personal grading rules; the initial regression network is tested using the sample test set; and the initial regression network that satisfies a preset error value is output as a first grading determination model; the initial regression network is preferably a deep neural network model.

[0128] S30: When the data category is enterprise data, a second economic value feature and a life cycle feature are obtained according to the enterprise data, and a corresponding enterprise classification rule is determined based on the second economic value feature and the life cycle feature. The enterprise classification rule at least includes a life cycle rule and a property rule.

[0129] It can be understood that the method for obtaining the second economic value feature based on the enterprise data can be the same as the method for obtaining the first economic value feature in the above step 20, except that there are some differences in the data. For example, the economic type at least includes financial type, social type and consumption type. Exemplarily, personal data related to the financial type can be income records, expenditure records, investment records, bank account information, etc., and personal data related to the social type can be social media activities, contact networks, social interaction records, etc. Then, relative to the enterprise, the enterprise data related to the financial type can be a balance sheet, an income statement, a cash flow statement, a tax return, a tax compliance record, etc., and the enterprise data related to the social type can be advertising records, social media marketing activities, marketing strategies, internal corporate emails, meeting minutes, etc. The specific calculation method can be the same, so this embodiment does not go into too much detail.

[0130] Methods for obtaining life cycle characteristics based on enterprise data include:

[0131] Preset the life cycle stage, determine the number of elements in the enterprise data that are in the life cycle stage, and use the number of elements in the life cycle stage as the life cycle feature. The life cycle stage at least includes the data storage stage, data usage stage, data transmission stage and data release stage.

[0132] In the above description, each element in the enterprise data refers to each piece of data in the enterprise data. Then, the method for determining the number of each element in the enterprise data in the life cycle stage includes:

[0133] Traverse each element in the enterprise data, determine whether it is in the data storage stage according to the storage time of each element, determine whether it is in the data storage stage according to the access time of each element, determine whether it is in the data transmission stage according to the transmission path of each element, and determine whether it is in the data release stage according to the release time of each element.

[0134] The method for determining the corresponding enterprise classification rules based on the second economic value characteristics and life cycle characteristics includes:

[0135] The second economic value characteristics and the life cycle characteristics are input into the pre-built second classification determination model to obtain the enterprise classification rules.

[0136] Similarly, the second grading determination model in this embodiment is constructed in the same manner as the first grading determination model in the above step S20, except that the input data is changed to the second economic value characteristics and life cycle characteristics, and the output data is changed to the enterprise grading rules.

[0137] S40: When the data category is public data, a leakage feature and a third economic value feature are obtained according to the public data, and a corresponding public classification rule is determined according to the leakage feature and the third economic value feature, where the public classification rule at least includes a leakage rule and a property rule;

[0138] It can be understood that the method for obtaining the third economic value feature based on public data can be the same as the method for obtaining the first economic value feature in the above step 20, except that there are some differences in the data. For example, the economic type includes at least financial type, social type and consumption type. For example, personal data related to the financial type can be income records, expenditure records, investment records, bank account information, etc., and personal data related to the social type can be social media activities, contact networks, social interaction records, etc. Then, with respect to public institutions, public data related to the financial type can be the annual fiscal budget of the government or public institution, expenditure details of various departments, project grant records, tax revenue data of public institutions, etc., and public data related to the social type can be public opinions and consultation records collected by public institutions through various channels, content published by public institutions on social media platforms, interaction records, etc. The specific calculation method can be the same, so this embodiment does not go into too much detail.

[0139] Methods for obtaining leakage characteristics based on public data include:

[0140] Obtain relevant sensitive data of public data, calculate the leakage risk coefficient based on the relevant sensitive data, and use the leakage risk coefficient as the leakage feature. The relevant sensitive data at least includes the number of access anomalies, the number of transmission path nodes, the number of target entities, and the historical leakage recovery cost;

[0141] It can be understood that the number of abnormal access times refers to the number of times public data is accessed at abnormal times or abnormal geographical locations, the number of transmission path nodes refers to the number of nodes involved in public data transmission, the number of target entities refers to the number of individuals or institutions affected by public data, and the leakage recovery cost refers to the cost of recovery after public data leakage. Taking the number of abnormal access times and the number of target entities as examples, frequent abnormal access indicates that public data may be at risk, and the larger the number of target entities, the more people or organizations will be affected after the data leakage, and the more serious the consequences will be.

[0142] Methods for calculating leakage risk factors based on relevant sensitive data include:

[0143]

[0144] Where LRC is the leakage risk coefficient, NAS is the number of access anomalies, NTP is the number of transmission path nodes, NTE is the number of target entities, cosh(·) is the hyperbolic cosine function, sinh[·] is the hyperbolic sine function, and 1! is the factorial of 1.

[0145] In this embodiment, taking the number of access anomalies as an example, the greater the number of access anomalies, the greater the corresponding leakage risk coefficient. Since the leakage risk coefficient is used as a leakage feature, the leakage feature is positively correlated with the leakage risk level of public data. Therefore, the greater the leakage feature, the more it is necessary to classify the public data according to the leakage rules.

[0146] Example 3

[0147] like Figure 4 As shown, this embodiment discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, the above-mentioned data security management and control method based on the data classification and grading method is implemented.

[0148] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters, weights and thresholds in the formula are set by technicians in this field according to actual conditions.

[0149] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center through a wired network or a wireless network. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD) or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0150] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0151] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0152] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0153] 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 on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0154] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0155] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0156] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A data security management and control system based on a data classification and grading method, characterized in that: include: Classification module: used to obtain the data to be processed, input the data to be processed into the pre-built data classification model, and obtain the data category, which includes personal data, enterprise data and public data; The first classification module is used for obtaining the first economic value feature and the legal association feature according to the personal data when the data category is personal data, and determining the corresponding personal classification rules based on the first economic value feature and the legal association feature, wherein the personal classification rules at least include responsibility rules and property rules; The second classification module is used for obtaining the second economic value characteristics and life cycle characteristics according to the enterprise data when the data category is enterprise data, and determining the corresponding enterprise classification rules based on the second economic value characteristics and life cycle characteristics. The enterprise classification rules at least include life cycle rules and property rules; The third classification module: when the data category is public data, it obtains the leakage characteristics and the third economic value characteristics according to the public data, and determines the corresponding public classification rules according to the leakage characteristics and the third economic value characteristics. The public classification rules at least include leakage rules and property rules.

2. According to claim 1, a data security management and control system based on data classification and grading method is characterized in that: The method for constructing the data classification model includes: Get r groups of data, where r is a positive integer greater than 1, and the data include historical data to be processed and historical data categories. Take the historical data to be processed and the historical data categories as sample sets, divide the sample sets into training sets and test sets, build a classifier, take the historical data to be processed in the training set as input data, take the historical data categories in the training set as output data, train the classifier to obtain an initial classifier, test the initial classifier with the test set, and output a classifier that meets the preset accuracy as a data classification model.

3. A data security management and control system based on data classification and grading method according to claim 2, characterized in that: The method for obtaining the first economic value feature according to personal data includes: The personal data is clustered according to the economic type to generate a first clustering result, and the first value characteristic is determined based on the first clustering result and the corresponding first influencing factor. The personal data is clustered according to the time period to generate a second clustering result, and the second value characteristic is determined based on the second clustering result and the corresponding second influencing factor. The personal data is clustered according to the data usage scope to generate a third clustering result, and the third value characteristic is determined based on the third clustering result and the corresponding third influencing factor. The first economic value characteristic is calculated based on the first value characteristic, the second value characteristic and the third value characteristic.

4. A data security management and control system based on a data classification and grading method according to claim 3, characterized in that: The method for calculating the first economic value feature based on the first value feature, the second value feature and the third value feature includes: In the formula, EVF is the first economic value feature, FVF is the first value feature, SVF is the second value feature, TVF is the third value feature, cot -1 (·) is the inverse cotangent function, cos -1 [·] is the arccosine function, sin -1 [·] is the inverse sine function, exp(·) is the exponential function with e as the base, log 10 (·) is a logarithmic function with base 10, and K is a constant greater than 0.

5. According to claim 1, a data security management and control system based on a data classification and grading method is characterized in that: The method of obtaining legal association characteristics based on personal data includes: According to the pre-built keyword extraction model, N target keywords are extracted from personal data. The target knowledge graph is constructed with the N target keywords as nodes and the relationships between the N target keywords as edges. The complexity of the target knowledge graph is calculated and the complexity is used as the legal association feature.

6. A data security management and control system based on data classification and grading method according to claim 5, characterized in that: The method for calculating the complexity of the target knowledge graph includes: Where Cplx is the complexity, N is the number of nodes in the target knowledge graph, and E is the number of edges in the target knowledge graph.

7. A data security management and control system based on a data classification and grading method according to claim 4, characterized in that: The method for determining corresponding individual classification rules based on the first economic value feature and the legal association feature includes: The first economic value feature and the legal association feature are input into a pre-constructed first grading determination model to obtain a personal grading rule. The method for constructing the first hierarchical determination model includes: A sample data set is obtained, wherein the sample data set includes a historical first economic value feature, a historical legal association feature, and a historical personal grading rule; the sample data set is divided into a sample training set and a sample test set; a regression network is constructed; the historical first economic value feature and the historical legal association feature in the sample training set are used as input data of the regression network; the historical personal grading rules in the sample training set are used as output data of the regression network; the regression network is trained to obtain an initial regression network for predicting real-time personal grading rules; the initial regression network is tested using the sample test set; and the initial regression network that satisfies a preset error value is output as a first grading determination model.

8. The data security management and control system based on the data classification and grading method according to claim 1 is characterized in that: The method for acquiring life cycle characteristics based on enterprise data includes: Preset the life cycle stage, determine the number of elements in the enterprise data that are in the life cycle stage, and use the number of elements in the life cycle stage as the life cycle feature. The life cycle stage at least includes the data storage stage, data usage stage, data transmission stage and data release stage.

9. A data security management and control system based on a data classification and grading method according to claim 2, characterized in that: The method for obtaining leakage characteristics according to public data includes: Obtain relevant sensitive data of public data, calculate the leakage risk coefficient based on the relevant sensitive data, and use the leakage risk coefficient as the leakage feature. The relevant sensitive data at least includes the number of access exceptions, the number of transmission path nodes, the number of target entities, and the historical leakage recovery cost.

10. A data security management and control method based on a data classification and grading method, which is implemented based on a data security management and control system based on a data classification and grading method as described in any one of claims 1 to 9, characterized in that: include: Obtaining the data to be processed, inputting the data to be processed into a pre-built data classification model to obtain data categories, where the data categories include personal data, enterprise data, and public data; When the data category is personal data, a first economic value feature and a legal association feature are obtained according to the personal data, and a corresponding personal classification rule is determined based on the first economic value feature and the legal association feature. The personal classification rule at least includes a responsibility rule and a property rule; When the data category is enterprise data, the second economic value feature and the life cycle feature are obtained according to the enterprise data, and the corresponding enterprise classification rules are determined based on the second economic value feature and the life cycle feature. The enterprise classification rules at least include life cycle rules and property rules; When the data category is public data, the leakage characteristics and the third economic value characteristics are obtained according to the public data, and the corresponding public classification rules are determined according to the leakage characteristics and the third economic value characteristics. The public classification rules at least include leakage rules and property rules.

Images