Enterprise information security management method and system based on cloud computing

Abstract

The invention relates to the technical field of information management, and discloses an enterprise information security management method and system based on cloud computing. Firstly, an initial enterprise information security management matrix is screened based on correlation analysis, then enterprise information security management data is obtained, and an improved support vector machine algorithm is used for data cleaning; secondly, calculating the weight of an enterprise information security management index, constructing a whitening weight function matrix and a gray evaluation matrix according to the weight, and calculating an enterprise information security risk value; constructing an initial deep learning algorithm model, obtaining a deep learning algorithm prediction model after training, and outputting an enterprise information security risk prediction value; finally, the enterprise information security state is judged according to the enterprise information security risk prediction value, and enterprise information security management is completed. By analyzing and processing the enterprise information security data, the purpose of enterprise information security management is achieved, and the method is objective and accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of information management, and in particular to a cloud computing-based enterprise information security management method and system. Background Art

[0002] Chinese patent CN117009589B discloses a safety inspection management method and system based on multimodal intelligent services. The method specifically includes: first, multimodal processing of standard documents through multimodal intelligent services to obtain a structured inspection standard list; second, using the inspection items corresponding to the inspection standards in the structured inspection standard list to establish an inspection item library, and obtain the full-link safety management data generated by the inspection items in the closed-loop management process; then, based on the inspection item library and the full-link safety management data, a safety production knowledge graph is obtained; finally, based on the safety production knowledge graph, the data in the safety inspection management process is input into the database to obtain a multi-dimensional and full-chain safety production knowledge graph to complete the safety inspection management. This invention directly obtains management data without preprocessing the data, and the results are biased.

[0003] Traditional enterprise information security management methods usually manage redundant and complex data, which is not conducive to improving the level of enterprise information security management under big data conditions, nor can it guarantee the security and stability of enterprise information in a security risk environment. At the same time, traditional enterprise information security management methods do not use high-tech technologies such as machine learning, and urgently need to improve various enterprise information security management strategies. Summary of the invention

[0004] In view of the problems in the related technology, the present invention provides an enterprise information security management method and system based on cloud computing to overcome the above-mentioned technical problems existing in the existing related technology.

[0005] In order to solve the above technical problems, the present invention is achieved through the following technical solutions:

[0006] The present invention is a cloud computing-based enterprise information security management method, comprising the following steps:

[0007] S1. Obtain enterprise information security management indicators, generate an initial enterprise information security management matrix, screen the initial enterprise information security management matrix based on correlation analysis, obtain an enterprise information security management set, generate an enterprise information security management data matrix, and perform data cleaning based on an improved support vector machine algorithm to obtain a processed enterprise information security management data matrix;

[0008] S2. Calculate the weights of the enterprise information security management indicators in the processed enterprise information security management data matrix to obtain a comprehensive weight set corresponding to the enterprise information security management indicators, construct a whitening weight function matrix and a gray evaluation matrix according to the comprehensive weight set, and calculate the enterprise information security risk value;

[0009] S3, constructing an initial deep learning algorithm model, retraining the initial deep learning algorithm model, obtaining a deep learning algorithm prediction model, labeling the enterprise information security risk value, and inputting it into the deep learning algorithm prediction model to obtain an enterprise information security risk prediction value;

[0010] S4. Determine the enterprise information security status based on the enterprise information security risk prediction value, formulate a risk avoidance plan, implement risk control, and complete enterprise information security management.

[0011] The invention performs correlation analysis on the initial enterprise information security management matrix, calculates the correlation coefficient, and uses the correlation coefficient and threshold to compare and then screen the enterprise information security management indicators; and uses an improved support vector machine algorithm to introduce the AdaBoost weak feature classifier to identify whether the enterprise information security management data is abnormal; the correlation analysis method can screen out indicators with strong correlation, which can not only highlight the dominant indicators, but also reduce unnecessary data operations; the improved support vector machine algorithm can realize the operation of multiple types of support vector machine algorithms, improve classification performance, and have good recognition effect; secondly, calculate the weights of enterprise information security management indicators, and construct A whitening weight function matrix and a grey evaluation matrix are constructed to calculate the enterprise information security risk value; this method introduces the D number theory to improve the fuzzy preference relationship, which can effectively reduce the influence of subjective factors and improve the accuracy of the results; then the initial deep learning algorithm model is constructed, and the initial deep learning algorithm model is trained to obtain the deep learning algorithm prediction model, and the enterprise information security risk prediction value is output; the neural network has a good fitting effect on most samples, and the prediction result has high accuracy, fast convergence speed and strong applicability; finally, the enterprise information security status is judged according to the enterprise information security risk prediction value, a risk avoidance plan is formulated, risk control is achieved, and enterprise information security management is completed.

[0012] Preferably, the S1 comprises the following steps:

[0013] S11. Obtain enterprise information security management indicators. Each enterprise information security management indicator is divided into j sub-indicators to generate an initial enterprise information security management matrix A as follows:

[0014]

[0015] Among them, a ij Represents the jth sub-indicator under the i-th enterprise information security management indicator;

[0016] The enterprise information security management indicators in the initial enterprise information security management matrix are screened using a correlation analysis method to obtain an enterprise information security management matrix. The specific steps are as follows:

[0017] S111. Sample the sub-indicators under the enterprise information security management index to obtain the sub-indicator sampling values. Select any sub-indicator under the enterprise information security management index, record them as the first sub-indicator and the second sub-indicator, calculate the average value of the first sub-indicator sampling value and the average value of the second sub-indicator sampling value, and use the correlation coefficient to express it. The calculation formula is as follows:

[0018]

[0019] Among them, β represents the correlation coefficient, Indicates the sampling value of the first sub-indicator of the i1th order, Indicates the i1th second sub-indicator sampling value, Indicates the average value of the first sub-indicator sampling value, represents the average value of the second sub-indicator sampling value, j1 represents the number of the first sub-indicator and the second sub-indicator sampling value, i1=1,2,3,...,j1;

[0020] S112, sequentially calculating the correlation coefficients of all sub-indicators under the enterprise information security management indicator, generating a correlation coefficient matrix, screening according to the correlation coefficients in the correlation coefficient matrix, setting a first coefficient threshold and a second coefficient threshold; when the correlation coefficient in the correlation coefficient matrix is ​​greater than or equal to the first coefficient threshold, deleting one sub-indicator corresponding to the correlation coefficient; when the correlation coefficient in the correlation coefficient matrix is ​​less than the first coefficient threshold and greater than or equal to the second coefficient threshold, deleting two sub-indicators corresponding to the correlation coefficient; until all correlation coefficients in the correlation coefficient matrix are screened, obtaining the screened sub-indicators,

[0021] Generate enterprise information security management collection;

[0022] S12, set the enterprise information security management set to include i' screened sub-indicators, quantify the screened sub-indicators in the enterprise information security management system to obtain enterprise information security management data, and collect j' enterprise information security management data under the screened sub-indicators to generate the enterprise information security management data matrix A1 as follows:

[0023]

[0024] Among them, a i′j′ represents the j′th enterprise information security management data under the i′th filtered sub-indicator;

[0025] Use the improved support vector machine algorithm and introduce the AdaBoost weak feature classifier to train the AdaBoost weak feature classifier to obtain the final feature classifier. The specific steps are as follows:

[0026] S121, reacquire the enterprise information security management data under the newly screened sub-indicators to obtain a sample data set, and divide the sample data set into a sample data training set and a sample data test set; introduce an AdaBoost weak feature classifier, set a number of AdaBoost weak feature classifiers in the support vector machine model, set the maximum number of iterations of the AdaBoost weak feature classifier, set the weight of the first AdaBoost weak feature classifier to α1, input the enterprise information security management data training set and the first AdaBoost weak feature classifier weight into the first AdaBoost weak feature classifier, and when the current number of iterations reaches the maximum number of iterations, stop iterating to obtain a weight α2; then input the enterprise information security management data training set and the weight α2 into the second AdaBoost weak feature classifier until all AdaBoost weak feature classifiers output weights to obtain a final weight;

[0027] S122, combining all weights to obtain a trained feature classifier, setting the kernel function of the support vector machine to a radial basis function; inputting the sample data test set into the trained feature classifier, setting the error threshold to ω, and obtaining the final feature classifier when the output result error is less than the error threshold; otherwise, adjusting the support vector machine parameters until the output result error is less than the error threshold;

[0028] S13. Divide the enterprise information security management data matrix into i′ enterprise information security management data sets according to the screened sub-indicators, input the i′ enterprise information security management data sets into the final feature classifier, output the classification result 1 to indicate that the corresponding enterprise information security management data is normal, output the classification result 0 to indicate that the corresponding enterprise information security management data is abnormal, delete the abnormal data, and use the missing forest algorithm to fill in the data to obtain the processed enterprise information security management data matrix.

[0029] The invention performs correlation analysis on the initial enterprise information security management matrix, calculates the correlation coefficient, and uses the correlation coefficient and threshold to compare and then screen the enterprise information security management indicators. It can screen out indicators with strong correlation, highlight the dominant indicators, and reduce unnecessary data operations; and uses an improved support vector machine algorithm. By introducing the AdaBoost weak feature classifier, it can realize the operation of multi-class support vector machine algorithms, improve classification performance, and achieve good recognition effects.

[0030] Preferably, S2 comprises the following steps:

[0031] S21, assigning security management evaluation values ​​to the screened sub-indicators in the processed enterprise information security management data matrix to obtain a security management evaluation value set, constructing a D-number security management evaluation matrix according to the security management evaluation value set, and using an integrated representation to convert the D-number security management evaluation matrix into a definite number security management evaluation matrix;

[0032] Calculate the occurrence probability of the determined number of safety management assessment data in the determined number of safety management assessment matrix, generate a probability matrix, calculate the safety management assessment probability of the sub-indicators screened in the probability matrix, and sort the safety management assessment probabilities in descending order to obtain the final D number safety management assessment matrix;

[0033] S22, set the weight vector set of the sub-indicators selected in the final D-number safety management evaluation matrix to B = {b1, b2, b3, ..., b m}, where b m It represents the weight vector of the mth selected sub-indicator in the final D-number security management evaluation matrix. The weight vector set of the selected sub-indicators in the processed enterprise information security management data matrix is ​​B1 = {b′1, b′2, b′3, ..., b′ m}, where b′ m represents the weight vector of the mth screened sub-indicator in the processed enterprise information security management data matrix, then the comprehensive weight set is B2 = {ω1,ω2,ω3,...,ω m}, where ω m represents the mth comprehensive weight, and ω m =b m b′ m ;

[0034] S23, constructing a whitened weight function matrix and a grey evaluation matrix according to the comprehensive weight set, and outputting the enterprise information security risk value, the specific steps are as follows:

[0035] S231, setting the whitening weight function to be an upper limit whitening weight function, a moderate whitening weight function and a lower limit whitening weight function, respectively, assigning different thresholds to the whitening weight function, respectively, to obtain a threshold set, performing whitening weight processing on the security management evaluation value set according to the threshold set, and converting the security management evaluation value set into a whitening weight function matrix A2 as follows:

[0036]

[0037] Among them, a mj″ represents the whitened weight corresponding to the j″th enterprise information security management data under the mth filtered sub-indicator;

[0038] S232, set the gray class set to B3 = {b″1, b″2, b″3, ..., b″ m′}, where b″ m′ represents the m′th gray class, solves the gray statistics corresponding to the whitening weights in the whitening weight function matrix according to the gray class set, calculates the average gray statistics of the sub-indicators after screening in the whitening weight function matrix, records it as the gray evaluation weight, obtains the gray evaluation weights corresponding to the sub-indicators after screening in the whitening weight function matrix, and forms a gray evaluation weight set; maps the gray evaluation weights in the gray evaluation weight set to the gray class set, and constructs a gray evaluation matrix;

[0039] S233, according to the comprehensive weight set and the grey evaluation matrix, a comprehensive evaluation vector is obtained, and the calculation formula is as follows:

[0040] B′=B2·B4;

[0041] Among them, B′ represents the comprehensive evaluation vector, and B4 represents the grey evaluation matrix;

[0042] The enterprise information security risk level is set to obtain an enterprise information security risk level vector, and the comprehensive evaluation vector and the trace of the enterprise information security risk level vector are multiplied to obtain the enterprise information security risk value.

[0043] The invention calculates the weights of enterprise information security management indicators, improves the fuzzy preference relationship by introducing the D-number theory, and constructs a whitened weight function matrix and a grey evaluation matrix, which can effectively reduce the influence of subjective factors and obtain an enterprise information security risk value with high accuracy.

[0044] Preferably, S3 comprises the following steps:

[0045] S31. Obtain the enterprise information security management data matrix of previous years, obtain the processed enterprise information security management data matrix of previous years after data cleaning, calculate the enterprise information security risk value, and form an enterprise information security risk value set; construct a multi-layer neural network, construct an initial deep learning algorithm model, and obtain a deep learning algorithm prediction model after training. The specific steps are as follows:

[0046] S311, set the multilayer neural network to include an input layer, a fully connected layer and an output layer, the number of neurons in the input layer is δ1, the fully connected layer includes δ2 neurons, set the input weight matrix and the output weight matrix, wherein the input weight matrix connects the input layer, the output weight matrix connects the output layer, and the number of units in the output layer is 1; introduce the nonlinear property σ as the activation function, the loss function is the mean square error, and construct the initial deep learning algorithm model;

[0047] S312, labeling the enterprise information security risk values ​​in the enterprise information security risk value set, generating a risk value sample set, dividing the risk value sample set into a risk value sample training set and a risk value sample test set, inputting the risk value sample training set into the initial deep learning algorithm model, setting the maximum number of iterations to C, the current number of iterations to c, and when the current number of iterations reaches the maximum number of iterations, stopping the iteration to obtain a trained deep learning algorithm model;

[0048] S313, input the risk value sample test set into the trained deep learning algorithm model, set the accuracy threshold, and when the output result accuracy is greater than the accuracy threshold, stop the iteration to obtain the deep learning algorithm prediction model; otherwise, continuously adjust the weight to minimize the loss function until the output result accuracy is greater than the accuracy threshold;

[0049] S32. Convert the enterprise information security risk value into a label, input it into the deep learning algorithm prediction model, output the enterprise information security risk prediction score, convert the enterprise information security risk prediction score into an enterprise information security risk prediction value, and complete the enterprise information security risk prediction.

[0050] This invention constructs an initial deep learning algorithm model and then trains the initial deep learning algorithm model to obtain a deep learning algorithm prediction model. The neural network has a good fitting effect on most data samples, has a fast convergence speed and strong applicability, and outputs enterprise information security risk prediction values ​​with high accuracy.

[0051] Preferably, S4 comprises the following steps:

[0052] S41, setting the risk threshold to ξ, when the enterprise information security risk prediction value is less than the risk threshold, the enterprise information security status is in a good state; otherwise, the enterprise information security is in an abnormal state, and the enterprise information security risk assessment is completed;

[0053] S42. For potential risks of enterprise information security, collect and screen them through risk supervision, obtain risk screening results, and establish a risk database and cloud server. Store the enterprise information security risk prediction values ​​and risk screening results in the risk database, use the cloud server to identify the risk screening results, formulate risk avoidance plans, implement risk control, and complete enterprise information security management.

[0054] The present invention also discloses a system of an enterprise information security management method based on cloud computing, which specifically includes: an enterprise information security management indicator screening module, an enterprise information security management data cleaning module, an enterprise information security risk value calculation module, an enterprise information security risk prediction module and an enterprise information security management module;

[0055] The enterprise information security management indicator screening module is used to screen the initial enterprise information security management indicators based on correlation analysis;

[0056] The enterprise information security management data cleaning module is used to perform data cleaning using an improved support vector machine algorithm;

[0057] The enterprise information security risk value calculation module is used to calculate the enterprise information security risk value in combination with the comprehensive weight of the enterprise information security management index;

[0058] The enterprise information security risk prediction module is used to predict enterprise information security risks using a deep learning algorithm prediction model;

[0059] The enterprise information security management module is used to formulate a risk avoidance plan for risk control according to the enterprise information security status.

[0060] The present invention has the following beneficial effects:

[0061] 1. The invention performs correlation analysis through the initial enterprise information security management matrix, calculates the correlation coefficient, and uses the correlation coefficient and the threshold to compare and screen the enterprise information security management indicators. It can screen out indicators with strong correlation, highlight the leading indicators, and reduce unnecessary data calculations;

[0062] 2. The invention uses an improved support vector machine algorithm and introduces the AdaBoost weak feature classifier to implement the operation of a multi-class support vector machine algorithm, improve classification performance, and achieve good recognition effect;

[0063] 3. The invention calculates the weights of enterprise information security management indicators, introduces the D-number theory to improve the fuzzy preference relationship, and constructs a whitened weight function matrix and a gray evaluation matrix, which can effectively reduce the influence of subjective factors and obtain a high accuracy of the enterprise information security risk value;

[0064] 4. The invention constructs an initial deep learning algorithm model and then trains the initial deep learning algorithm model to obtain a deep learning algorithm prediction model. The neural network has a good fitting effect on most data samples, has a fast convergence speed and strong applicability, and the output enterprise information security risk prediction value has high accuracy.

[0065] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, they can also obtain drawings based on these drawings without paying creative work.

[0067] Figure 1 A schematic diagram of the process of enterprise information security management by a cloud computing-based enterprise information security management system provided by the present invention. DETAILED DESCRIPTION

[0068] 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.

[0069] In the description of the present invention, it is necessary to understand that the terms "opening", "upper", "lower", "top", "middle", "inside" and the like indicating orientation or positional relationship are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the invention.

[0070] Example 1

[0071] Please refer to Figure 1 The present invention is a cloud computing-based enterprise information security management method, comprising the following steps:

[0072] S1. Obtain enterprise information security management indicators, generate an initial enterprise information security management matrix, screen the initial enterprise information security management matrix based on correlation analysis, obtain an enterprise information security management set, generate an enterprise information security management data matrix, and perform data cleaning based on an improved support vector machine algorithm to obtain a processed enterprise information security management data matrix;

[0073] The S1 comprises the following steps:

[0074] S11. Obtain enterprise information security management indicators. Each enterprise information security management indicator is divided into j sub-indicators to generate an initial enterprise information security management matrix A as follows:

[0075]

[0076] Among them, a ij Represents the jth sub-indicator under the i-th enterprise information security management indicator;

[0077] The enterprise information security management indicators in the initial enterprise information security management matrix are screened using a correlation analysis method to obtain an enterprise information security management matrix. The specific steps are as follows:

[0078] S111. Sample the sub-indicators under the enterprise information security management index to obtain the sub-indicator sampling values. Select any sub-indicator under the enterprise information security management index, record them as the first sub-indicator and the second sub-indicator, calculate the average value of the first sub-indicator sampling value and the average value of the second sub-indicator sampling value, and use the correlation coefficient to express it. The calculation formula is as follows:

[0079]

[0080] Among them, β represents the correlation coefficient, Indicates the sampling value of the first sub-indicator of the i1th order, Indicates the i1th second sub-indicator sampling value, Indicates the average value of the first sub-indicator sampling value, represents the average value of the second sub-indicator sampling value, j1 represents the number of the first sub-indicator and the second sub-indicator sampling value, i1=1,2,3,...,j1;

[0081] S112, sequentially calculating the correlation coefficients of all sub-indicators under the enterprise information security management indicator, generating a correlation coefficient matrix, screening according to the correlation coefficients in the correlation coefficient matrix, setting a first coefficient threshold and a second coefficient threshold; when the correlation coefficient in the correlation coefficient matrix is ​​greater than or equal to the first coefficient threshold, deleting one sub-indicator corresponding to the correlation coefficient, and when the correlation coefficient in the correlation coefficient matrix is ​​less than the first coefficient threshold and greater than or equal to the second coefficient threshold, deleting two sub-indicators corresponding to the correlation coefficient; until all correlation coefficients in the correlation coefficient matrix are screened, the screened sub-indicators are obtained, and an enterprise information security management set is generated;

[0082] S12, set the enterprise information security management set to include i' screened sub-indicators, quantify the screened sub-indicators in the enterprise information security management system to obtain enterprise information security management data, and collect j' enterprise information security management data under the screened sub-indicators to generate the enterprise information security management data matrix A1 as follows:

[0083]

[0084] Among them, a i′j′ represents the j′th enterprise information security management data under the i′th filtered sub-indicator;

[0085] Use the improved support vector machine algorithm and introduce the AdaBoost weak feature classifier to train the AdaBoost weak feature classifier to obtain the final feature classifier. The specific steps are as follows:

[0086] S121, reacquire the enterprise information security management data under the newly screened sub-indicators to obtain a sample data set, and divide the sample data set into a sample data training set and a sample data test set; introduce an AdaBoost weak feature classifier, set a number of AdaBoost weak feature classifiers in the support vector machine model, set the maximum number of iterations of the AdaBoost weak feature classifier, set the weight of the first AdaBoost weak feature classifier to α1, input the enterprise information security management data training set and the first AdaBoost weak feature classifier weight into the first AdaBoost weak feature classifier, and when the current number of iterations reaches the maximum number of iterations, stop iterating to obtain a weight α2; then input the enterprise information security management data training set and the weight α2 into the second AdaBoost weak feature classifier until all AdaBoost weak feature classifiers output weights to obtain a final weight;

[0087] S122, combining all weights to obtain a trained feature classifier, setting the kernel function of the support vector machine to a radial basis function; inputting the sample data test set into the trained feature classifier, setting the error threshold to ω, and obtaining the final feature classifier when the output result error is less than the error threshold; otherwise, adjusting the support vector machine parameters until the output result error is less than the error threshold;

[0088] S13, dividing the enterprise information security management data matrix into i' enterprise information security management data sets according to the screened sub-indicators, inputting the i' enterprise information security management data sets into the final feature classifier, outputting a classification result of 1 to indicate that the corresponding enterprise information security management data is normal, outputting a classification result of 0 to indicate that the corresponding enterprise information security management data is abnormal, deleting the abnormal data, and using the missing forest algorithm to fill in the data to obtain a processed enterprise information security management data matrix;

[0089] S2. Calculate the weights of the enterprise information security management indicators in the processed enterprise information security management data matrix to obtain a comprehensive weight set corresponding to the enterprise information security management indicators, construct a whitening weight function matrix and a gray evaluation matrix according to the comprehensive weight set, and calculate the enterprise information security risk value;

[0090] The S2 comprises the following steps:

[0091] S21, assigning security management evaluation values ​​to the screened sub-indicators in the processed enterprise information security management data matrix to obtain a security management evaluation value set, constructing a D-number security management evaluation matrix according to the security management evaluation value set, and using an integrated representation to convert the D-number security management evaluation matrix into a definite number security management evaluation matrix;

[0092] Calculate the occurrence probability of the determined number of safety management assessment data in the determined number of safety management assessment matrix, generate a probability matrix, calculate the safety management assessment probability of the sub-indicators screened in the probability matrix, and sort the safety management assessment probabilities in descending order to obtain the final D number safety management assessment matrix;

[0093] S22, set the weight vector set of the sub-indicators selected in the final D-number safety management evaluation matrix to B = {b1, b2, b3, ..., b m}, where b m It represents the weight vector of the mth selected sub-indicator in the final D-number security management evaluation matrix. The weight vector set of the selected sub-indicators in the processed enterprise information security management data matrix is ​​B1 = {b′1, b′2, b′3, ..., b′ m}, where b′ m represents the weight vector of the mth screened sub-indicator in the processed enterprise information security management data matrix, then the comprehensive weight set is B2 = {ω1,ω2,ω3,...,ω m}, where ω m represents the mth comprehensive weight, and ω m =b m b′ m ;

[0094] S23, constructing a whitened weight function matrix and a grey evaluation matrix according to the comprehensive weight set, and outputting the enterprise information security risk value, the specific steps are as follows:

[0095] S231, setting the whitening weight function to be an upper limit whitening weight function, a moderate whitening weight function and a lower limit whitening weight function, respectively, assigning different thresholds to the whitening weight function, respectively, to obtain a threshold set, performing whitening weight processing on the security management evaluation value set according to the threshold set, and converting the security management evaluation value set into a whitening weight function matrix A2 as follows:

[0096]

[0097] Among them, a mj″ represents the whitened weight corresponding to the j″th enterprise information security management data under the mth filtered sub-indicator;

[0098] S232, set the gray class set to B3 = {b″1, b″2, b″3, ..., b″ m′}, where b″ m′ represents the m′th gray class, solves the gray statistics corresponding to the whitening weights in the whitening weight function matrix according to the gray class set, calculates the average gray statistics of the sub-indicators after screening in the whitening weight function matrix, records it as the gray evaluation weight, obtains the gray evaluation weights corresponding to the sub-indicators after screening in the whitening weight function matrix, and forms a gray evaluation weight set; maps the gray evaluation weights in the gray evaluation weight set to the gray class set, and constructs a gray evaluation matrix;

[0099] S233, according to the comprehensive weight set and the grey evaluation matrix, a comprehensive evaluation vector is obtained, and the calculation formula is as follows:

[0100] B′=B2·B4;

[0101] Among them, B′ represents the comprehensive evaluation vector, and B4 represents the grey evaluation matrix;

[0102] Setting the enterprise information security risk level, obtaining the enterprise information security risk level vector, performing multiplication operation on the comprehensive evaluation vector and the trace of the enterprise information security risk level vector, and obtaining the enterprise information security risk value;

[0103] S3, constructing an initial deep learning algorithm model, retraining the initial deep learning algorithm model, obtaining a deep learning algorithm prediction model, labeling the enterprise information security risk value, and inputting it into the deep learning algorithm prediction model to obtain an enterprise information security risk prediction value;

[0104] The S3 comprises the following steps:

[0105] S31. Obtain the enterprise information security management data matrix of previous years, obtain the processed enterprise information security management data matrix of previous years after data cleaning, calculate the enterprise information security risk value, and form an enterprise information security risk value set; construct a multi-layer neural network, construct an initial deep learning algorithm model, and obtain a deep learning algorithm prediction model after training. The specific steps are as follows:

[0106] S311, set the multilayer neural network to include an input layer, a fully connected layer and an output layer, the number of neurons in the input layer is δ1, the fully connected layer includes δ2 neurons, set the input weight matrix and the output weight matrix, wherein the input weight matrix connects the input layer, the output weight matrix connects the output layer, and the number of units in the output layer is 1; introduce the nonlinear property σ as the activation function, the loss function is the mean square error, and construct the initial deep learning algorithm model;

[0107] S312, labeling the enterprise information security risk values ​​in the enterprise information security risk value set, generating a risk value sample set, dividing the risk value sample set into a risk value sample training set and a risk value sample test set, inputting the risk value sample training set into the initial deep learning algorithm model, setting the maximum number of iterations to C, the current number of iterations to c, and when the current number of iterations reaches the maximum number of iterations, stopping the iteration to obtain a trained deep learning algorithm model;

[0108] S313, input the risk value sample test set into the trained deep learning algorithm model, set the accuracy threshold, and when the output result accuracy is greater than the accuracy threshold, stop the iteration to obtain the deep learning algorithm prediction model; otherwise, continuously adjust the weight to minimize the loss function until the output result accuracy is greater than the accuracy threshold;

[0109] S32, converting the enterprise information security risk value into a label, and then inputting it into the deep learning algorithm prediction model, outputting the enterprise information security risk prediction score, converting the enterprise information security risk prediction score into the enterprise information security risk prediction value, and completing the enterprise information security risk prediction;

[0110] S4. Determine the enterprise information security status according to the enterprise information security risk prediction value, formulate a risk avoidance plan, implement risk control, and complete enterprise information security management;

[0111] The S4 comprises the following steps:

[0112] S41, setting the risk threshold to ξ, when the enterprise information security risk prediction value is less than the risk threshold, the enterprise information security status is in a good state; otherwise, the enterprise information security is in an abnormal state, and the enterprise information security risk assessment is completed;

[0113] S42. For potential risks of enterprise information security, collect and screen them through risk supervision, obtain risk screening results, and establish a risk database and cloud server. Store the enterprise information security risk prediction values ​​and risk screening results in the risk database, use the cloud server to identify the risk screening results, formulate risk avoidance plans, implement risk control, and complete enterprise information security management.

[0114] Example 2

[0115] The present invention also discloses a system of an enterprise information security management method based on cloud computing, which specifically includes: an enterprise information security management indicator screening module, an enterprise information security management data cleaning module, an enterprise information security risk value calculation module, an enterprise information security risk prediction module and an enterprise information security management module;

[0116] The enterprise information security management indicator screening module is used to screen the initial enterprise information security management indicators based on correlation analysis;

[0117] The enterprise information security management data cleaning module is used to perform data cleaning using an improved support vector machine algorithm;

[0118] The enterprise information security risk value calculation module is used to calculate the enterprise information security risk value in combination with the comprehensive weight of the enterprise information security management index;

[0119] The enterprise information security risk prediction module is used to predict enterprise information security risks using a deep learning algorithm prediction model;

[0120] The enterprise information security management module is used to formulate a risk avoidance plan for risk control according to the enterprise information security status.

[0121] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0122] The preferred embodiments of the invention disclosed above are only used to help explain the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the invention, so that those skilled in the art can understand and use the invention well.

Claims

1. A cloud computing-based enterprise information security management method, characterized in that: The steps include: S1. Obtain enterprise information security management indicators, generate an initial enterprise information security management matrix, screen the initial enterprise information security management matrix based on correlation analysis, obtain an enterprise information security management set, generate an enterprise information security management data matrix, and perform data cleaning based on an improved support vector machine algorithm to obtain a processed enterprise information security management data matrix; S2. Calculate the weights of the enterprise information security management indicators in the processed enterprise information security management data matrix to obtain a comprehensive weight set corresponding to the enterprise information security management indicators, construct a whitening weight function matrix and a gray evaluation matrix according to the comprehensive weight set, and calculate the enterprise information security risk value; S3, constructing an initial deep learning algorithm model, retraining the initial deep learning algorithm model, obtaining a deep learning algorithm prediction model, labeling the enterprise information security risk value, and inputting it into the deep learning algorithm prediction model to obtain an enterprise information security risk prediction value; S4. Determine the enterprise information security status based on the enterprise information security risk prediction value, formulate a risk avoidance plan, implement risk control, and complete enterprise information security management.

2. The enterprise information security management method based on cloud computing according to claim 1, characterized in that: The S1 comprises the following steps: S11, obtaining enterprise information security management indicators, each enterprise information security management indicator is divided into j sub-indicators, and an initial enterprise information security management matrix is ​​generated; the enterprise information security management indicators in the initial enterprise information security management matrix are screened using a correlation analysis method to obtain an enterprise information security management matrix; S12, setting the enterprise information security management set to include i′ screened sub-indicators, quantifying the screened sub-indicators to obtain enterprise information security management data, and generating an enterprise information security management data matrix; using an improved support vector machine algorithm, and introducing an AdaBoost weak feature classifier, training the AdaBoost weak feature classifier to obtain a final feature classifier; S13. Input the enterprise information security management data matrix into the final feature classifier, output the classification result 1 to indicate that the corresponding enterprise information security management data is normal, output the classification result 0 to indicate that the corresponding enterprise information security management data is abnormal, delete the abnormal data, and perform data filling to obtain the processed enterprise information security management data matrix.

3. The enterprise information security management method based on cloud computing according to claim 2 is characterized in that: The S11 comprises the following steps: S111. Sampling data of sub-indicators under the enterprise information security management indicator to obtain sub-indicator sampling values, and calculating correlation coefficients of the sub-indicator sampling values; S112. Calculate the correlation coefficients of all sub-indicators under the enterprise information security management indicator in sequence to generate a correlation coefficient matrix, filter the correlation coefficients in the correlation coefficient matrix to obtain filtered sub-indicators, and generate an enterprise information security management set.

4. The enterprise information security management method based on cloud computing according to claim 3 is characterized in that: The improved support vector machine algorithm is used, and the AdaBoost weak feature classifier is introduced. Training the AdaBoost weak feature classifier to obtain the final feature classifier includes the following steps: Re-acquire the enterprise information security management data under the newly screened sub-indicators to obtain a sample data set, and divide the sample data set into a sample data training set and a sample data test set; introduce an AdaBoost weak feature classifier, set a number of AdaBoost weak feature classifiers in the support vector machine model, set the maximum number of iterations of the AdaBoost weak feature classifier, set the weight of the first AdaBoost weak feature classifier to α1, input the enterprise information security management data training set and the first AdaBoost weak feature classifier weight into the first AdaBoost weak feature classifier, and when the current number of iterations reaches the maximum number of iterations, stop the iteration to obtain the weight α2; then input the enterprise information security management data training set and the weight α2 into the second AdaBoost weak feature classifier until all AdaBoost weak feature classifiers output weights to obtain the final weight; Combine all weights to obtain a trained feature classifier, and set the kernel function of the support vector machine to the radial basis function; Input the sample data test set into the trained feature classifier, set the error threshold to ω, and when the output result error is less than the error threshold, obtain the final feature classifier; Otherwise, adjust the support vector machine parameters until the output result error is less than the error threshold.

5. The enterprise information security management method based on cloud computing according to claim 4 is characterized in that: The S2 comprises the following steps: S21. Based on the D-number theory, convert the processed enterprise information security management data matrix into a final D-number security management evaluation matrix; S22, calculating the weight vectors of the selected sub-indicators in the final D-number safety management evaluation matrix to obtain a comprehensive weight set; S23. Construct a whitened weight function matrix and a grey evaluation matrix according to the comprehensive weight set, and output the enterprise information security risk value.

6. The enterprise information security management method based on cloud computing according to claim 5, characterized in that: The S3 comprises the following steps: S31. Obtain the enterprise information security management data matrix of previous years, and after data cleaning, obtain the processed enterprise information security management data matrix of previous years, calculate the enterprise information security risk value, and form an enterprise information security risk value set; Construct a multi-layer neural network, construct an initial deep learning algorithm model, and obtain a deep learning algorithm prediction model after training; S32. Convert the enterprise information security risk value into a label, input it into the deep learning algorithm prediction model, output the enterprise information security risk prediction score, convert the enterprise information security risk prediction score into an enterprise information security risk prediction value, and complete the enterprise information security risk prediction.

7. The enterprise information security management method based on cloud computing according to claim 6 is characterized in that: The steps of constructing a multi-layer neural network, constructing an initial deep learning algorithm model, and obtaining a deep learning algorithm prediction model after training include the following steps: The multilayer neural network is set to include an input layer, a fully connected layer and an output layer, the number of neurons in the input layer is δ1, the fully connected layer includes δ2 neurons, the input weight matrix and the output weight matrix are set, wherein the input weight matrix connects the input layer, the output weight matrix connects the output layer, and the number of units in the output layer is 1; the nonlinear property σ is introduced as the activation function, the loss function is the mean square error, and the initial deep learning algorithm model is constructed; Label the enterprise information security risk values ​​in the enterprise information security risk value set to generate a risk value sample set, divide the risk value sample set into a risk value sample training set and a risk value sample test set, input the risk value sample training set into the initial deep learning algorithm model, set the maximum number of iterations to C, the current number of iterations to c, and when the current number of iterations reaches the maximum number of iterations, stop iterating to obtain a trained deep learning algorithm model; Then input the risk value sample test set into the trained deep learning algorithm model, set the accuracy threshold, and when the output result accuracy is greater than the accuracy threshold, stop the iteration to obtain the deep learning algorithm prediction model; otherwise, continuously adjust the weight to minimize the loss function until the output result accuracy is greater than the accuracy threshold.

8. The enterprise information security management method based on cloud computing according to claim 7, characterized in that: The S4 comprises the following steps: S41, setting the risk threshold to ξ, when the enterprise information security risk prediction value is less than the risk threshold, the enterprise information security status is in a good state; Otherwise, the enterprise information security is in an abnormal state and the enterprise information security risk assessment is completed; S42. For potential risks of enterprise information security, collect and screen them through risk supervision, obtain risk screening results, and establish a risk database and cloud server. Store the enterprise information security risk prediction values ​​and risk screening results in the risk database, use the cloud server to identify the risk screening results, formulate risk avoidance plans, implement risk control, and complete enterprise information security management.

9. A system for implementing the enterprise information security management method based on cloud computing as claimed in any one of claims 1 to 8, characterized in that: Specifically include: Enterprise information security management indicator screening module, enterprise information security management data cleaning module, enterprise information security risk value calculation module, enterprise information security risk prediction module and enterprise information security management module; The enterprise information security management indicator screening module is used to screen the initial enterprise information security management indicators based on correlation analysis; The enterprise information security management data cleaning module is used to perform data cleaning using an improved support vector machine algorithm; The enterprise information security risk value calculation module is used to calculate the enterprise information security risk value in combination with the comprehensive weight of the enterprise information security management index; The enterprise information security risk prediction module is used to predict enterprise information security risks using a deep learning algorithm prediction model; The enterprise information security management module is used to formulate a risk avoidance plan for risk control according to the enterprise information security status.

Images