A method, system, and storage medium for predicting data security risks on a cloud platform.

Abstract

This invention discloses a method, system, and storage medium for predicting data security risks on a cloud platform, relating to the field of data security technology. The method includes: obtaining at least one data storage node and at least one data storage path of the cloud platform; obtaining the data importance level of the data storage node; backing up the data on the cloud platform, with the backup data stored in a blockchain; obtaining at least one type of network attack and calculating the proportion of each type; calculating the average impact factor of each network attack type on the cloud platform; calculating the overall impact factor of at least one network attack on the cloud platform; and using the overall impact factor as the predicted level of data security risks on the cloud platform. By setting up a security classification module, an anomaly acquisition module, a factor calculation module, and a risk prediction module, and considering the different situations of various network attacks and their different impacts on the data, the obtained prediction results have high reference value.

Description

Technical Field

[0001] This invention relates to the field of data security technology, specifically to a method, system, and storage medium for predicting data security risks on a cloud platform. Background Technology

[0002] With the rapid development of the digital economy and the advancement of digital transformation of traditional businesses, data security has become the most urgent and fundamental security issue in the digital economy era. At present, there are a large number of complex data processing activities. How to systematically, comprehensively and effectively build and improve data security protection capabilities, achieve a balance between data economy, data value and data security, perceive and monitor data theft, and predict potential data security risks in the future to avoid data breaches are issues of great concern to the industry and enterprises.

[0003] Data security risk prediction involves complex cyberattacks and complex data security issues. Existing risk prediction methods do not take into account all kinds of situations, resulting in predictions that are not specific enough and have low reference value. Summary of the Invention

[0004] To address the aforementioned technical issues, this paper provides a cloud platform data security risk prediction method, system, and storage medium. This technical solution resolves the problems mentioned in the background section regarding data security risk prediction involving complex network attacks and complex data security issues. Existing risk prediction methods do not comprehensively consider various situations, resulting in predictions that are not specific enough and have low reference value.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A method for predicting data security risks on a cloud platform includes:

[0007] Obtain the data storage structure of the cloud platform, and based on the data storage structure, obtain at least one data storage node and at least one data storage path of the cloud platform.

[0008] Obtain the data importance level of the data storage nodes;

[0009] Data on the cloud platform is backed up and stored in the blockchain. The blockchain is configured with access verification, which includes facial recognition and iris recognition. If both facial recognition and iris recognition are passed, access permission to access data in the blockchain is obtained. If both facial recognition and iris recognition are not passed, access permission to access data in the blockchain cannot be obtained.

[0010] The system identifies at least one cyberattack against the cloud platform, classifies the cyberattack to obtain at least one cyberattack category, and calculates the percentage of each cyberattack category.

[0011] To determine the extent of data leakage following a network attack on a cloud platform, within the network attack category;

[0012] Obtain the degree of anomaly in the storage path after a network attack on the cloud platform, within the network attack category;

[0013] Obtain the percentage of false data following attacks on cloud platforms within the network attack category;

[0014] The average impact factor of network attack categories on cloud platform attacks was calculated.

[0015] The overall impact factor of at least one network attack on cloud platform attacks was calculated.

[0016] The overall impact factor is used as the prediction level of cloud platform data security risk. The larger the overall impact factor, the greater the data security risk of the cloud platform.

[0017] Preferably, obtaining the data importance level of the data storage node includes the following steps:

[0018] Construct a neural network model, use the neural network model to identify keywords in the data of all data storage nodes, and classify the keywords into levels. The higher the importance of the keyword, the higher the keyword level.

[0019] Based on the hierarchical division of keywords, the data of all data storage nodes is hierarchically divided to obtain the importance level of each data.

[0020] The importance level of each data point in a single data storage node is obtained by weighting and summing the importance levels of each data point in that single data storage node. The weighting coefficients represent the proportion of each data point in the single data storage node.

[0021] Preferably, obtaining at least one network attack category and calculating the proportion of network attack categories includes the following steps:

[0022] Build a neural network model and use it to classify at least one network attack.

[0023] Network attacks that are categorized into similar types are grouped into network attack categories;

[0024] Count the total number of attacks involving at least one network attack;

[0025] The number of network attacks within each category is counted as the attack component count;

[0026] Dividing the number of attack components by the total number of attacks yields the percentage of each type of network attack.

[0027] Preferably, the process of obtaining the extent of data leakage following a network attack on the cloud platform, within the network attack category, includes the following steps:

[0028] In the category of network attacks, after a network attack targets a cloud platform, the amount of leaked data output from the data storage node to outside the cloud platform is obtained.

[0029] The weighted leak amount is obtained by multiplying the data importance level of the data storage node by the amount of leaked data.

[0030] The weighted sum of the leakage values ​​across all data storage nodes is used as the degree of data leakage following a network attack on the cloud platform, within the network attack category.

[0031] Preferably, obtaining the degree of abnormality of the storage path after a network attack on the cloud platform, within the network attack category, includes the following steps:

[0032] In the category of network attacks, after a network attack targets a cloud platform, data is exchanged using the data storage path.

[0033] If there is an anomaly in the data interaction, the data storage path will be used as the abnormal storage path; otherwise, no action will be taken.

[0034] The number of data storage paths that are used as abnormal storage paths is counted to obtain the number of abnormalities;

[0035] Count the number of at least one data storage path to obtain the total number of paths;

[0036] The number of anomalies divided by the total number of paths is used as the degree of storage path anomaly after a network attack on the cloud platform, within the network attack category.

[0037] Preferably, obtaining the percentage of false data following a network attack on the cloud platform within the network attack category includes the following steps:

[0038] In the category of network attacks, after a network attack targets a cloud platform, the attacker obtains the post-attack data from the data storage nodes.

[0039] Access verification is set through the blockchain to retrieve the feature backup data corresponding to the data storage node in the blockchain.

[0040] Data obtained after an attack on a data storage node that differs from the characteristic backup data is used as part of the fake data;

[0041] By summing up the partial spurious data from all data storage nodes, the total spurious data is obtained.

[0042] Obtain the total data volume across all data storage nodes;

[0043] The total number of fake data items is divided by the total amount of data to determine the percentage of fake data items resulting from a network attack on the cloud platform within the network attack category.

[0044] Preferably, the calculation of the average impact factor of network attack categories on cloud platform attacks includes the following steps:

[0045] The average data leakage rate is obtained by summing up the data leakage rates of all network attacks in the network attack category after attacking the cloud platform and averaging them.

[0046] The average storage path anomaly level is obtained by summing the anomalies of all network attacks in the network attack category after attacking the cloud platform and averaging them.

[0047] The average percentage of false data is obtained by summing the percentages of false data after all network attacks in the network attack category on the cloud platform and averaging them.

[0048] The average impact factor is obtained by summing the average data leakage level, the average storage path anomaly level, and the average proportion of fake data.

[0049] Preferably, calculating the overall impact factor of at least one network attack on the cloud platform includes the following steps:

[0050] Obtain at least one network attack category;

[0051] Multiply the average impact factor of each cyberattack category by the proportion of each cyberattack category to obtain the average weighted impact factor of each cyberattack category.

[0052] By summing the average weighted impact factors of at least one network attack category, we obtain the overall impact factor of at least one network attack on the cloud platform.

[0053] A cloud platform data security risk prediction system, used to implement the above-mentioned cloud platform data security risk prediction method, includes:

[0054] The data acquisition module acquires at least one data storage node and at least one data storage path of the cloud platform.

[0055] The data classification module obtains the data importance level of the data storage nodes;

[0056] The data backup module backs up the data on the cloud platform, and the backup data is stored in the blockchain.

[0057] The security classification module acquires at least one network attack that the cloud platform has been subjected to, classifies the at least one network attack, and obtains at least one network attack category.

[0058] An anomaly acquisition module, which acquires the degree of data leakage after a network attack on the cloud platform in the network attack category, acquires the degree of storage path anomaly after a network attack on the cloud platform in the network attack category, and acquires the proportion of false data after a network attack on the cloud platform in the network attack category.

[0059] The factor calculation module calculates the average impact factor of network attack categories on cloud platform attacks, and calculates the overall impact factor of at least one network attack on cloud platform attacks.

[0060] The risk prediction module uses the overall impact factor as the prediction level for cloud platform data security risks.

[0061] A storage medium having a computer-readable program stored thereon, wherein the computer-readable program, when invoked, executes the aforementioned cloud platform data security risk prediction method.

[0062] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0063] By setting up modules for data grading, security classification, anomaly detection, factor calculation, and risk prediction, this study uses the impact of existing cyberattacks on the cloud platform to infer the platform's resilience. During prediction, cyberattacks are categorized, and the importance of cloud platform data is classified. Then, based on data leakage, data changes, and path corruption, the risk level of each cyberattack on the cloud platform is assessed. Finally, the overall risk impact of all cyberattacks on the cloud platform is obtained. The prediction process considers the different scenarios of various cyberattacks and their varying impacts on the data, resulting in a high degree of consistency between the predictions and the cloud platform, making it a valuable reference. Attached Figure Description

[0064] Figure 1 This is a schematic diagram of the cloud platform data security risk prediction method of the present invention;

[0065] Figure 2 This is a schematic diagram illustrating the process of obtaining the data importance level of a data storage node according to the present invention;

[0066] Figure 3 A schematic diagram illustrating the process of calculating the proportion of network attack categories to obtain at least one network attack category in this invention;

[0067] Figure 4 This is a schematic diagram illustrating the process of obtaining the degree of data leakage after a network attack on a cloud platform, as described in this invention.

[0068] Figure 5This is a schematic diagram illustrating the process of obtaining the degree of storage path anomaly after a network attack on a cloud platform, as described in this invention.

[0069] Figure 6 This is a schematic diagram illustrating the process of obtaining the percentage of false data following a network attack on a cloud platform within the network attack category, as described in this invention.

[0070] Figure 7 This is a schematic diagram illustrating the process of calculating the average impact factor of network attack categories on cloud platform attacks in this invention.

[0071] Figure 8 This is a schematic diagram illustrating the process of calculating the overall impact factor of at least one network attack on a cloud platform in accordance with the present invention. Detailed Implementation

[0072] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0073] Reference Figure 1 As shown, a method for predicting data security risks on a cloud platform includes:

[0074] Obtain the data storage structure of the cloud platform, and based on the data storage structure, obtain at least one data storage node and at least one data storage path of the cloud platform.

[0075] Obtain the data importance level of the data storage nodes;

[0076] Data on the cloud platform is backed up and stored in the blockchain. The blockchain is configured with access verification, which includes facial recognition and iris recognition. If both facial recognition and iris recognition are passed, access permission to access data in the blockchain is obtained. If both facial recognition and iris recognition are not passed, access permission to access data in the blockchain cannot be obtained.

[0077] The system identifies at least one cyberattack against the cloud platform, classifies the cyberattack to obtain at least one cyberattack category, and calculates the percentage of each cyberattack category.

[0078] To determine the extent of data leakage following a network attack on a cloud platform, within the network attack category;

[0079] Obtain the degree of anomaly in the storage path after a network attack on the cloud platform, within the network attack category;

[0080] Obtain the percentage of false data following attacks on cloud platforms within the network attack category;

[0081] The average impact factor of network attack categories on cloud platform attacks was calculated.

[0082] The overall impact factor of at least one network attack on cloud platform attacks was calculated.

[0083] The overall impact factor is used as the prediction level of cloud platform data security risk. The larger the overall impact factor, the greater the data security risk of the cloud platform.

[0084] Reference Figure 2 As shown, obtaining the data importance level of a data storage node includes the following steps:

[0085] Construct a neural network model, use the neural network model to identify keywords in the data of all data storage nodes, and classify the keywords into levels. The higher the importance of the keyword, the higher the keyword level.

[0086] Based on the hierarchical division of keywords, the data of all data storage nodes is hierarchically divided to obtain the importance level of each data.

[0087] The importance level of each data point in a single data storage node is obtained by weighting and summing the importance levels of each data point in that single data storage node, where the weighting coefficients represent the proportion of each data point in that single data storage node.

[0088] The purpose of obtaining the importance level of data storage nodes is to predict data anomalies after a network attack. Under the same data anomaly conditions, the more important the data, the higher the risk. Therefore, it is necessary to classify the importance of the data. If the data is not differentiated, the accuracy of the prediction results will be reduced.

[0089] Reference Figure 3 As shown, obtaining at least one network attack category and calculating the proportion of network attack categories involves the following steps:

[0090] Build a neural network model and use it to classify at least one network attack.

[0091] Network attacks that are categorized into similar types are grouped into network attack categories;

[0092] Count the total number of attacks involving at least one network attack;

[0093] The number of network attacks within each category is counted as the attack component count;

[0094] Dividing the number of attack components by the total number of attacks yields the percentage of each type of network attack.

[0095] Cyberattacks can be categorized based on their effects, and their codes can be mapped to these categories. A standard neural network model is then constructed to identify and classify cyberattack codes, assigning each attack to its corresponding category. The purpose of calculating the percentage of each cyberattack category is to weight the impact of different cyberattack categories on cloud platform data when assessing their influence, thereby synthesizing the risk outcomes for all cyberattack categories.

[0096] Reference Figure 4 As shown, obtaining the extent of data leakage after a network attack on a cloud platform, within the network attack category, includes the following steps:

[0097] In the category of network attacks, after a network attack targets a cloud platform, the amount of leaked data output from the data storage node to outside the cloud platform is obtained.

[0098] The weighted leak amount is obtained by multiplying the data importance level of the data storage node by the amount of leaked data.

[0099] The weighted leakage values ​​of all data storage nodes are summed up to represent the degree of data leakage after a network attack on the cloud platform, which falls under the category of network attacks.

[0100] When data is leaked, data is inevitably output outside the cloud platform. Therefore, the amount of output data is considered as the amount of leaked data. The risk of data leakage is related to the importance of the data; the higher the importance, the greater the degree of leakage.

[0101] Reference Figure 5 As shown, obtaining the degree of anomaly in the storage path after a network attack on the cloud platform, within the network attack category, includes the following steps:

[0102] In the category of network attacks, after a network attack targets a cloud platform, data is exchanged using the data storage path.

[0103] If there is an anomaly in the data interaction, the data storage path will be used as the abnormal storage path; otherwise, no action will be taken.

[0104] The number of data storage paths that are used as abnormal storage paths is counted to obtain the number of abnormalities;

[0105] Count the number of at least one data storage path to obtain the total number of paths;

[0106] The number of anomalies divided by the total number of paths is used as the degree of storage path anomaly after a network attack on the cloud platform, within the network attack category.

[0107] Data storage paths may be damaged after a cyberattack, resulting in the inability to transmit data and affecting the normal use of the cloud platform. Therefore, it is necessary to assess the degree of abnormality of the storage path.

[0108] Reference Figure 6 As shown, obtaining the percentage of fake data following a network attack on a cloud platform includes the following steps:

[0109] In the category of network attacks, after a network attack targets a cloud platform, the attacker obtains the post-attack data from the data storage nodes.

[0110] Access verification is set through the blockchain to retrieve the feature backup data corresponding to the data storage node in the blockchain.

[0111] Data obtained after an attack on a data storage node that differs from the characteristic backup data is used as part of the fake data;

[0112] By summing up the partial spurious data from all data storage nodes, the total spurious data is obtained.

[0113] Obtain the total data volume across all data storage nodes;

[0114] The total number of fake data is divided by the total amount of data to determine the proportion of fake data following a network attack on a cloud platform within the network attack category.

[0115] During cyberattacks, not only is data leakage possible, but there is also the possibility of injecting false data into the cloud platform, thus polluting the data within the cloud platform. Therefore, it is necessary to compare the amount of false data in order to estimate the impact of each cyberattack on the cloud platform.

[0116] Reference Figure 7 As shown, the calculation of the average impact factor of network attack categories on cloud platform attacks includes the following steps:

[0117] The average data leakage rate is obtained by summing up the data leakage rates of all network attacks in the network attack category after attacking the cloud platform and averaging them.

[0118] The average storage path anomaly level is obtained by summing the anomalies of all network attacks in the network attack category after attacking the cloud platform and averaging them.

[0119] The average percentage of false data is obtained by summing the percentages of false data after all network attacks in the network attack category on the cloud platform and averaging them.

[0120] The average impact factor is obtained by summing the average data leakage level, the average storage path anomaly level, and the average proportion of fake data.

[0121] When calculating the average impact factor of network attack categories on cloud platform attacks, the degree of data leakage, the degree of storage path anomaly, and the proportion of fake data are combined to take into account the influence of multiple factors, making the prediction more comprehensive and the results more accurate.

[0122] Since the network environment is stable for a period of time and the attacks it receives are generally similar, the prediction results generated by this method can predict the risks of the cloud platform for a period of time. However, when the preset time is exceeded, the prediction results need to be updated using this method, and the updated results are used to predict the risks. If the predicted risks are within an acceptable range, the cloud platform will not be upgraded or modified. If the predicted risks are not within an acceptable range, the cloud platform will be upgraded or modified.

[0123] Reference Figure 8 As shown, calculating the overall impact factor of at least one network attack on a cloud platform includes the following steps:

[0124] Obtain at least one network attack category;

[0125] Multiply the average impact factor of each cyberattack category by the proportion of each cyberattack category to obtain the average weighted impact factor of each cyberattack category.

[0126] By summing the average weighted impact factors of at least one network attack category, we obtain the overall impact factor of at least one network attack on the cloud platform.

[0127] A cloud platform data security risk prediction system, used to implement the above-mentioned cloud platform data security risk prediction method, includes:

[0128] The data acquisition module acquires at least one data storage node and at least one data storage path of the cloud platform.

[0129] The data classification module obtains the data importance level of the data storage nodes;

[0130] The data backup module backs up the data on the cloud platform, and the backup data is stored in the blockchain.

[0131] The security classification module acquires at least one network attack that the cloud platform has been subjected to, classifies the at least one network attack, and obtains at least one network attack category.

[0132] An anomaly acquisition module, which acquires the degree of data leakage after a network attack on the cloud platform in the network attack category, acquires the degree of storage path anomaly after a network attack on the cloud platform in the network attack category, and acquires the proportion of false data after a network attack on the cloud platform in the network attack category.

[0133] The factor calculation module calculates the average impact factor of network attack categories on cloud platform attacks, and calculates the overall impact factor of at least one network attack on cloud platform attacks.

[0134] The risk prediction module uses the overall impact factor as the prediction level for cloud platform data security risks.

[0135] The working process of the cloud platform data security risk prediction system is as follows;

[0136] Step 1: The data acquisition module obtains the data storage structure of the cloud platform, and based on the data storage structure, obtains at least one data storage node and at least one data storage path of the cloud platform.

[0137] Step 2: The data grading module obtains the data importance level of the data storage nodes;

[0138] Step 3: The data backup module backs up the data on the cloud platform, and the backup data is stored in the blockchain;

[0139] Step 4: The security classification module obtains at least one network attack that the cloud platform has been subjected to, classifies the at least one network attack to obtain at least one network attack category, and calculates the proportion of network attack categories.

[0140] Step 5: The anomaly acquisition module obtains the degree of data leakage after the network attack on the cloud platform in the network attack category, the degree of storage path anomaly after the network attack on the cloud platform in the network attack category, and the proportion of false data after the network attack on the cloud platform in the network attack category.

[0141] Step Six: The factor calculation module calculates the average impact factor of network attack categories on cloud platform attacks, and calculates the overall impact factor of at least one network attack on cloud platform attacks.

[0142] Step 7: The risk prediction module uses the overall impact factor as the prediction level of cloud platform data security risk. The higher the overall impact factor, the greater the risk to cloud platform data security.

[0143] Furthermore, this solution also proposes a storage medium on which a computer-readable program is stored, which, when invoked, executes the aforementioned cloud platform data security risk prediction method.

[0144] It is understandable that the storage medium can be a magnetic medium, such as a floppy disk, hard disk, or magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid-state drive (SSD).

[0145] In summary, the advantages of this invention are as follows: By setting up a data classification module, a security classification module, an anomaly acquisition module, a factor calculation module, and a risk prediction module, the invention uses the impact of existing network attacks on the cloud platform to infer the cloud platform's risk resistance capability. During the prediction process, network attacks are classified, and the importance of the cloud platform's data is categorized. Then, based on the degree of data leakage, data change, and path corruption, the risk level of each network attack on the cloud platform is assessed. Finally, the risk impact of all network attacks on the cloud platform is comprehensively obtained. In the prediction process, the different situations of various network attacks and their different impacts on the data are taken into account, resulting in a high degree of matching between the prediction results and the cloud platform, and thus possessing high reference value.

[0146] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for predicting data security risks on a cloud platform, characterized in that, include: Obtain the data storage structure of the cloud platform, and based on the data storage structure, obtain at least one data storage node and at least one data storage path of the cloud platform. Obtain the data importance level of the data storage nodes; Data on the cloud platform is backed up and stored in the blockchain. The blockchain is configured with access verification, which includes facial recognition and iris recognition. If both facial recognition and iris recognition are passed, access permission to access data in the blockchain is obtained. If both facial recognition and iris recognition are not passed, access permission to access data in the blockchain cannot be obtained. The system identifies at least one cyberattack that has been launched against the cloud platform, classifies the cyberattack to obtain at least one cyberattack category, and calculates the percentage of each cyberattack category. To determine the extent of data leakage following a network attack on a cloud platform, within the network attack category; Obtain the degree of anomaly in the storage path after a network attack on the cloud platform, within the network attack category; Obtain the percentage of false data following attacks on cloud platforms within the category of network attacks; The average impact factor of network attack categories on cloud platform attacks was calculated. The overall impact factor of at least one network attack on cloud platform attacks was calculated. The overall impact factor is used as the prediction level of cloud platform data security risk. The larger the overall impact factor, the greater the data security risk of the cloud platform. The process of determining the extent of data leakage following a network attack on a cloud platform, within the network attack category, includes the following steps: In the category of network attacks, after a network attack targets a cloud platform, the amount of leaked data output from the data storage node to outside the cloud platform is obtained. The weighted leak amount is obtained by multiplying the data importance level of the data storage node by the amount of leaked data. The weighted leakage values ​​of all data storage nodes are summed up to represent the degree of data leakage after a network attack on the cloud platform, which falls under the category of network attacks. The process of determining the degree of abnormality of the storage path after a network attack on the cloud platform, within the network attack category, includes the following steps: In the category of network attacks, after a network attack targets a cloud platform, data interaction is performed using the data storage path. If there is an anomaly in the data interaction, the data storage path will be used as the abnormal storage path; otherwise, no action will be taken. The number of data storage paths that are used as abnormal storage paths is counted to obtain the number of abnormalities; Count the number of at least one data storage path to obtain the total number of paths; The number of anomalies divided by the total number of paths is used as the degree of storage path anomaly after a network attack on the cloud platform, within the network attack category. The calculation of the average impact factor of network attack categories on cloud platform attacks includes the following steps: The average data leakage rate is obtained by summing up the data leakage rates of all network attacks in the network attack category after attacking the cloud platform and averaging them. The average storage path anomaly level is obtained by summing the anomalies of all network attacks in the network attack category after attacking the cloud platform and averaging them. The average percentage of false data is obtained by summing the percentages of false data after all network attacks in the network attack category on the cloud platform and averaging them. The average impact factor is obtained by summing the average data leakage level, the average storage path anomaly level, and the average proportion of fake data. The calculation of at least one overall impact factor of a network attack on a cloud platform includes the following steps: Obtain at least one network attack category; Multiply the average impact factor of each cyberattack category by the proportion of each cyberattack category to obtain the average weighted impact factor of each cyberattack category. By summing the average weighted impact factors of at least one network attack category, we obtain the overall impact factor of at least one network attack on the cloud platform.

2. The cloud platform data security risk prediction method according to claim 1, characterized in that, The process of obtaining the data importance level of the data storage node includes the following steps: Construct a neural network model, use the neural network model to identify keywords in the data of all data storage nodes, and classify the keywords into levels. The higher the importance of the keyword, the higher the keyword level. Based on the hierarchical division of keywords, the data in all data storage nodes is hierarchically divided to obtain the importance level of each data. The importance level of each data point in a single data storage node is obtained by weighting and summing the importance levels of each data point in that single data storage node. The weighting coefficients represent the proportion of each data point in the single data storage node.

3. The cloud platform data security risk prediction method according to claim 2, characterized in that, Obtaining at least one network attack category and calculating the proportion of network attack categories includes the following steps: Build a neural network model and use it to classify at least one network attack. Network attacks that are categorized into similar types are grouped into network attack categories; Count the total number of attacks involving at least one network attack; The number of network attacks within each category is counted as the attack component count; Dividing the number of attack components by the total number of attacks yields the percentage of each type of network attack.

4. The cloud platform data security risk prediction method according to claim 3, characterized in that, The process of obtaining the percentage of false data following a network attack on a cloud platform within the network attack category includes the following steps: In the category of network attacks, after a network attack targets a cloud platform, the attacker obtains the post-attack data from the data storage nodes. Access verification is set through the blockchain to retrieve the feature backup data corresponding to the data storage node in the blockchain. Data obtained after an attack on a data storage node that differs from the characteristic backup data is used as part of the fake data; By summing up the partial spurious data from all data storage nodes, the total spurious data is obtained. Obtain the total data volume across all data storage nodes; The total number of fake data items is divided by the total amount of data to determine the percentage of fake data items resulting from a network attack on the cloud platform within the network attack category.

5. A cloud platform data security risk prediction system, used to implement the cloud platform data security risk prediction method as described in any one of claims 1-4, characterized in that, include: The data acquisition module acquires at least one data storage node and at least one data storage path of the cloud platform. A data classification module, which acquires the data importance level of data storage nodes; The data backup module backs up the data on the cloud platform, and the backup data is stored in the blockchain; The security classification module acquires at least one network attack that the cloud platform has been subjected to, classifies the at least one network attack, and obtains at least one network attack category. An anomaly acquisition module, which acquires the degree of data leakage after a network attack on the cloud platform in the network attack category, acquires the degree of storage path anomaly after a network attack on the cloud platform in the network attack category, and acquires the proportion of false data after a network attack on the cloud platform in the network attack category. The factor calculation module calculates the average impact factor of network attack categories on cloud platform attacks, and calculates the overall impact factor of at least one network attack on cloud platform attacks. The risk prediction module uses the overall impact factor as the prediction level for cloud platform data security risks.

6. A storage medium having a computer-readable program stored thereon, characterized in that, When the computer-readable program is invoked, it executes the cloud platform data security risk prediction method as described in any one of claims 1-4.

Images