Risk assessment method for enterprise cross-domain data

Abstract

The invention discloses a risk assessment method for enterprise cross-domain data. The method comprises the following steps: collecting original assessment data of an enterprise for data exchange and sharing among different departments; preprocessing the collected original evaluation data, wherein the preprocessing comprises the steps of carrying out standardization processing on numerical data and carrying out fuzzification processing on text description data; constructing a risk assessment model based on a Bayesian network, and determining a dependency relationship and a conditional probability between risk factors by introducing expert knowledge and historical data; calculating the posterior probability of each risk factor through a Bayesian network; an expert evaluation method is adopted to allocate a weight for each risk evaluation index to carry out risk-based scoring on each risk factor; and calculating a comprehensive risk score of the enterprise cross-domain data through a comprehensive risk scoring formula. According to the method, an enterprise can be helped to accurately evaluate and manage risks in a complex data environment, and the data security and compliance are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise data security, and more specifically, to a risk assessment method for enterprise cross-domain data. Background Art

[0002] In the current context of informatization and globalization, data exchange and sharing between enterprises and between different departments within an enterprise has become an important part of daily operations. However, in the process of managing and utilizing cross-domain data, enterprises face a variety of potential risks, including data leakage, loss of data integrity, and compliance issues. These risks may not only cause economic losses to enterprises, but also lead to legal disputes and reputation damage. Therefore, enterprises urgently need to effectively evaluate and manage the risks of cross-domain data. Traditional data risk assessment methods usually rely on static rules and manual judgment, which are difficult to cope with complex and dynamically changing data environments. In addition, there are complex dependencies and uncertainties between many risk factors, which are difficult to quantify and deeply analyze by traditional methods. In the field of enterprise cross-domain data risk identification and management, international and national standards have covered many aspects. For example, ISO / IEC 27001 and ISO / IEC 27002 standards provide requirements and practical guidelines for information security management systems (ISMS) to help enterprises establish, implement, maintain and continuously improve information security management. In addition, ISO 31000 provides principles, frameworks and processes for risk management, which are suitable for all types of organizations to systematically identify, evaluate and manage risks. Although some experts, scholars and corporate researchers have proposed several risk assessment methods, these methods still have limitations when dealing with the uncertainty problem in the cross-domain data risk of enterprises. Therefore, how to effectively identify and assess the risks of cross-domain data of enterprises, especially under uncertain conditions, has become a key and important topic. Summary of the invention

[0003] The present invention proposes a risk assessment method for enterprise cross-domain data, which fully considers fuzzy uncertainty and aims to solve the problem of insufficient accuracy in risk identification and assessment of enterprise cross-domain data in the prior art. Through the method of the present invention, enterprises can more accurately assess and manage potential risks in complex data environments, significantly improve data security and compliance, effectively protect the data security of enterprises and support them to perform robust risk management in a changing environment. To achieve the above purpose, the present invention provides the following technical solutions:

[0004] A risk assessment method for enterprise cross-domain data, characterized by comprising:

[0005] Collect original assessment data for data exchange and sharing among different departments of the enterprise;

[0006] Preprocessing the collected raw evaluation data, including: standardizing the numerical data and fuzzifying the text description data;

[0007] Construct a risk assessment model based on Bayesian networks to determine the dependencies and conditional probabilities between various risk factors through expert knowledge and historical data;

[0008] The posterior probability of each risk factor was calculated using a Bayesian network;

[0009] The expert evaluation method is used to assign weights to each risk assessment indicator and to grade each risk factor;

[0010] The comprehensive risk score of an enterprise’s cross-domain data is calculated using the comprehensive risk score formula.

[0011] Preferably, the standardization of numerical data includes:

[0012] By formula: , and standardization is performed, where represents the jth eigenvalue of the i-th original numerical data, represents the minimum value among the j-th eigenvalues, represents the maximum value of the j-th eigenvalue, Represents the jth eigenvalue of the i-th normalized numerical data.

[0013] Preferably, the text description data is fuzzified, including:

[0014] Defining fuzzy sets , where each Represents a fuzzy set, converts text description data into fuzzy linguistic variables, and uses fuzzy membership function Map text data to corresponding fuzzy sets and define membership functions as follows: ;

[0015] in, represents the scoring result, and x represents the evaluation value.

[0016] Preferably, the step of determining the dependency relationship and conditional probability between various risk assessment indicators through expert knowledge and historical data includes:

[0017] Based on the historical data of the enterprise, the initial dependency relationship between the nodes in the Bayesian network is determined, and each node represents a specific risk factor.

[0018] Preferably, the step of determining the dependencies and conditional probabilities between various risk assessment indicators through expert knowledge and historical data further includes:

[0019] Data related to various risk factors are collected from the actual operating data of the enterprise. Based on expert knowledge and historical data, dependencies between nodes are established, and causal structures and potential correlations between various risk factors are determined.

[0020] Preferably, the step of determining the dependencies and conditional probabilities between various risk assessment indicators through expert knowledge and historical data further includes:

[0021] According to the collected data and the results of expert evaluation, the conditional probability between each node is set, and the conditional probability calculation formula is: , calculate the conditional probability between each node, where represents the conditional probability of X occurring when Y occurs, and are the prior probabilities of events X and Y, respectively. is the conditional probability that Y will occur given X.

[0022] Preferably, the calculating the posterior probability of each risk factor by using a Bayesian network includes:

[0023] Set the prior probability of nodes based on historical data or expert opinions ;

[0024] Using Bayesian theorem, update the posterior probability of the node based on the observed data , where the calculation formula for the posterior probability is: , represents the observed evidence, is the probability of evidence E appearing when node X occurs, is the overall probability of evidence E appearing.

[0025] Preferably, the expert evaluation method is used to assign weights to each risk assessment indicator, including:

[0026] Use expert evaluation method to assign weights to each risk assessment indicator. The allocation is based on the impact of risk factors on the overall risk of the enterprise, the interrelationships between risk factors and the statistical analysis results of historical data.

[0027] Preferably, the step of calculating the comprehensive risk score of the enterprise's cross-domain data includes:

[0028] According to the calculation formula: , the comprehensive risk score is calculated, where represents the posterior probability of the i-th risk factor, represents the weight of the ith risk factor, and n is the total number of risk factors.

[0029] Preferably, the original assessment data includes: data leakage risk DLR, data integrity risk DIR, compliance risk CR, sensitive data exposure level SDEL, data transmission speed DTS, access frequency AF, user behavior abnormality UBA and data access permission level DAPL.

[0030] The present invention provides a risk assessment method for enterprise cross-domain data, which solves the problem of inaccurate risk identification and assessment of existing enterprise cross-domain data, can improve the data security of the enterprise, can help the enterprise accurately assess and manage risks in a complex data environment, and improve data security and compliance. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the specific embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below.

[0032] Figure 1 It is a schematic diagram of a risk assessment method for enterprise cross-domain data provided by the present invention.

[0033] Figure 2 It is a flow chart of enterprise cross-domain data risk assessment provided by an embodiment of the present invention.

[0034] Figure 3 It is a schematic diagram of the set enterprise cross-domain data risk assessment parameters at various levels provided by an embodiment of the present invention.

[0035] Figure 4 It is a Bayesian network topology diagram provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0036] In order to enable persons skilled in the art to better understand the solutions of the embodiments of the present invention, the embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings and implementation modes.

[0037] In response to the current problem of inaccurate risk identification and assessment of cross-basin data by enterprises, the present invention provides a risk assessment method for enterprise cross-domain data, which solves the problem of inaccurate risk identification and assessment of cross-basin data by existing enterprises, can improve the data security of enterprises, and can help enterprises accurately assess and manage risks in complex data environments, thereby improving data security and compliance.

[0038] like Figure 1 and Figure 2 As shown, a risk assessment method for enterprise cross-domain data includes:

[0039] S1: Collect original assessment data for data exchange and sharing among different departments of the enterprise.

[0040] S2: Preprocessing the collected original evaluation data, including: standardizing the numerical data and fuzzifying the text description data.

[0041] S3: Construct a risk assessment model based on Bayesian network and determine the dependencies and conditional probabilities among various risk factors by introducing expert knowledge and historical data.

[0042] S4: Calculate the posterior probability of each risk factor through the Bayesian network.

[0043] S5: Use expert evaluation method to assign weights to each risk assessment indicator and to grade each risk factor.

[0044] S6: Calculate the comprehensive risk score of the enterprise's cross-domain data through the comprehensive risk score formula.

[0045] Specifically, the original assessment data exchanged and shared between different departments of the enterprise, including numerical data and text description data, are collected, the numerical data is standardized to ensure the consistency and comparability of the data, and the text description data is fuzzified to effectively handle fuzzy and uncertain information. This multi-source data comprehensive processing method can improve the comprehensiveness and accuracy of data preprocessing. Based on the Bayesian network, the risk assessment model is constructed. By introducing expert knowledge and historical data, the dependencies and conditional probabilities between various risk assessment indicators are determined. The Bayesian network can effectively process and express the complex dependencies between risk factors, which improves the scientificity and rationality of risk assessment. The posterior probability of each risk factor is calculated by the Bayesian network, and the uncertainty of each factor is fully considered. The expert evaluation method is used to assign weights to each risk assessment indicator, which ensures the professionalism and rationality of the weight assignment and further enhances the credibility of the assessment results. By comprehensively considering each risk assessment indicator and its weight, the comprehensive risk score of the enterprise's cross-domain data is calculated. This scoring method can not only quantify and describe the complex dependencies between risk factors, but also provide an intuitive risk assessment indicator, which is convenient for enterprises to conduct risk management and decision-making. This method can accurately describe and quantify the uncertainty and complex dependencies between risk factors, effectively improving the security and compliance of enterprise data management. Through systematic risk assessment and early warning mechanisms, it helps enterprises to timely identify and prevent potential risks and ensure the security and compliance of data management.

[0046] In one embodiment, the process of risk assessment for enterprise cross-domain data is as follows:

[0047] Step 1: Collect the original assessment data for data exchange and sharing between different departments of the enterprise, such as Figure 3As shown in Table 1, the original assessment data includes data leakage risk DLR, data integrity risk DIR, compliance risk CR, sensitive data exposure level SDEL, data transmission speed DTS, access frequency AF, user behavior abnormality UBA and data access permission level DAPL; therefore, the enterprise cross-domain data risk assessment model has a total of 8 assessment parameters as shown in Table 1.

[0048]

[0049] Step 2: In order to eliminate the influence of different data dimensions and the difference between numerical data and text description data in enterprise cross-domain data risk assessment, the original assessment data is preprocessed, including standardization and fuzzification, as follows:

[0050] Step 2.1: Standardize the numerical data. The specific formula is as follows:

[0051] in, represents the jth eigenvalue of the i-th original numerical data, represents the minimum value among the j-th eigenvalues, represents the maximum value of the j-th eigenvalue, Represents the jth eigenvalue of the i-th normalized numerical data.

[0052] The original numerical data are shown in Table 2, and the numerical data after normalization are shown in Table 3.

[0053]

[0054] Step 2.2: Fuzzify the text description data and define the fuzzy set , where each Represents a fuzzy set, such as "low risk", "medium risk" and "high risk". For text description data, convert it into fuzzy linguistic variables and use fuzzy membership function Map text data to corresponding fuzzy sets and define membership functions as follows: ,

[0055] in, represents the scoring result, and x represents the evaluation value.

[0056] The original text description data is shown in Table 4, and the text description data after fuzzification is shown in Table 5.

[0057]

[0058] Step 3: Construct a risk assessment model based on Bayesian network and determine the dependencies and conditional probabilities among various risk assessment indicators through expert knowledge and historical data.

[0059] Step 3.1: Based on the historical data of the enterprise, determine the initial dependency relationship between the nodes (risk factors) in the Bayesian network. Each node represents a specific risk factor. Figure 4 .

[0060] Step 3.2: Collect data related to each risk factor from the actual operational data of the enterprise, establish dependencies between nodes based on expert knowledge and historical data, and determine the causal structure and potential correlation between each risk factor.

[0061] Step 3.3: Use Bayes' theorem and conditional probability calculation formula to build a conditional probability table (CPT). Let X and Y be two interdependent nodes. It represents the probability of X occurring when Y occurs. The calculation formula of conditional probability is:

[0062] in, is the conditional probability, and are the prior probabilities of events X and Y, respectively. is the probability of Y occurring when X occurs. Based on the collected data and the results of expert evaluation, the conditional probabilities between the nodes are set. The calculated conditional probabilities between the nodes are shown in Tables 6 to 8.

[0063]

[0064] Step 4: Calculate the posterior probability of each risk factor through the Bayesian network and set the prior probability of the node based on historical data or expert opinions , using Bayesian theorem, update the posterior probability of the node according to the observed data ,in represents the observed evidence. The formula for calculating the posterior probability is:

[0065] in, is the probability of evidence E appearing when node X occurs, is the overall probability of the occurrence of evidence E. The posterior probability of each risk factor calculated by the Bayesian network is as follows:

[0066] ;

[0067] ;

[0068] ;

[0069] ;

[0070] ;

[0071] ;

[0072] ;

[0073] ;

[0074] ;

[0075] Step 5: Use expert evaluation method to assign weights to each risk assessment indicator. The allocation is based on the impact of risk factors on the overall risk of the enterprise, the correlation between risk factors and the statistical analysis results of historical data. The weight allocated to each risk assessment indicator is shown in Table 9.

[0076]

[0077] Step 6: Calculate the comprehensive risk score of the enterprise's cross-domain data. The calculation formula for the comprehensive risk score is:

[0078] in, represents the posterior probability of the ith risk factor, represents the weight of the ith risk factor, and n is the total number of risk factors.

[0079] Through the above calculations, the posterior probabilities of DLR, DIR, and CR at the highest risk are 0.6, 0.2, and 0.3 respectively. The comprehensive risk score of the enterprise's cross-domain data is calculated as follows:

[0080] Therefore, enterprises whose data risk is assessed as medium should take the following measures: strengthen data encryption and access control to reduce the risk of data leakage; optimize data transmission speed and access frequency to ensure data integrity; strengthen monitoring of user behavior and permission management to ensure compliance.

[0081] It can be seen that the present invention provides a risk assessment method for enterprise cross-domain data, which solves the problem of inaccurate risk identification and assessment of existing enterprise cross-domain data, can improve the data security of the enterprise, can help the enterprise accurately assess and manage risks in a complex data environment, and improve data security and compliance.

[0082] The above describes in detail the structure, features and effects of the present invention based on the embodiments shown in the drawings. The above is only a preferred embodiment of the present invention, but the present invention is not limited to the scope of implementation shown in the drawings. Any changes made in accordance with the concept of the present invention, or modifications to equivalent embodiments with equivalent changes, which still do not exceed the spirit covered by the description and drawings, should be within the protection scope of the present invention.

Claims

1. A risk assessment method for enterprise cross-domain data, characterized in that: include: Collect original assessment data for data exchange and sharing among different departments of the enterprise; Preprocessing the collected raw evaluation data, including: standardizing the numerical data and fuzzifying the text description data; Construct a risk assessment model based on Bayesian network, and determine the dependency and conditional probability between various risk factors by introducing expert knowledge and historical data; The posterior probability of each risk factor was calculated using a Bayesian network; The expert evaluation method is used to assign weights to each risk assessment indicator and to grade each risk factor; The comprehensive risk score of an enterprise’s cross-domain data is calculated using the comprehensive risk score formula.

2. The risk assessment method for enterprise cross-domain data according to claim 1 is characterized in that: The standardization of the numerical data includes: By formula: , and standardization is performed, where represents the jth eigenvalue of the i-th original numerical data, represents the minimum value among the j-th eigenvalues, represents the maximum value of the j-th eigenvalue, Represents the jth eigenvalue of the i-th normalized numerical data.

3. The risk assessment method for enterprise cross-domain data according to claim 2 is characterized in that: The fuzzy processing of the text description data includes: Defining fuzzy sets , where each Represents a fuzzy set, converts text description data into fuzzy linguistic variables, and uses fuzzy membership function Map text data to corresponding fuzzy sets and define membership functions as follows: ； in, represents the scoring result, and x represents the evaluation value.

4. The risk assessment method for enterprise cross-domain data according to claim 3 is characterized in that: The dependency relationship and conditional probability between various risk assessment indicators are determined by expert knowledge and historical data, including: Based on the historical data of the enterprise, the initial dependency relationship between the nodes in the Bayesian network is determined, and each node represents a specific risk factor.

5. The risk assessment method for enterprise cross-domain data according to claim 4 is characterized in that: Determining the dependencies and conditional probabilities between various risk assessment indicators through expert knowledge and historical data also includes: Data related to various risk factors are collected from the actual operating data of the enterprise. Based on expert knowledge and historical data, dependencies between nodes are established, and causal structures and potential correlations between various risk factors are determined.

6. The risk assessment method for enterprise cross-domain data according to claim 5 is characterized in that: Determining the dependencies and conditional probabilities between various risk assessment indicators through expert knowledge and historical data also includes: According to the collected data and the results of expert evaluation, the conditional probability between each node is set, and the conditional probability calculation formula is: , calculate the conditional probability between each node, where represents the conditional probability of X occurring when Y occurs, and are the prior probabilities of events X and Y, respectively. is the conditional probability that Y will occur given X.

7. The risk assessment method for enterprise cross-domain data according to claim 6 is characterized in that: The posterior probability of each risk factor is calculated by using a Bayesian network, including: Set the prior probability of nodes based on historical data or expert opinions ; Using Bayesian theorem, update the posterior probability of the node based on the observed data , where the calculation formula for the posterior probability is: , represents the observed evidence, is the probability of evidence E appearing when node X occurs, is the overall probability of evidence E appearing.

8. The risk assessment method for enterprise cross-domain data according to claim 7, characterized in that: The expert evaluation method is used to assign weights to each risk assessment indicator, including: Use expert evaluation method to assign weights to each risk assessment indicator. The allocation is based on the impact of risk factors on the overall risk of the enterprise, the interrelationships between risk factors and the statistical analysis results of historical data.

9. The risk assessment method for enterprise cross-domain data according to claim 8, characterized in that: The calculation of the comprehensive risk score of the enterprise's cross-domain data includes: According to the calculation formula: , the comprehensive risk score is calculated, where represents the posterior probability of the i-th risk factor, represents the weight of the ith risk factor, and n is the total number of risk factors.

10. The risk assessment method for enterprise cross-domain data according to any one of claims 1 to 9, characterized in that: The original assessment data includes: data leakage risk DLR, data integrity risk DIR, compliance risk CR, sensitive data exposure level SDEL, data transmission speed DTS, access frequency AF, user behavior abnormality UBA and data access permission level DAPL.

Images