Mobile phone security module management system and method

By setting a detection window in the mobile phone security module and using deep learning models to analyze the sort code correlation, identifying and managing high-risk sort codes, the data leakage problem caused by the failure to effectively identify the sort code correlation in the existing technology is solved, and the data leakage problem caused by the failure to effectively identify the sort code correlation is improved. The security and stability of the system are prevented from large-scale data breaches and fraud.

CN118761069BActive Publication Date: 2024-11-15CHANGSHA LEILEIYUN NETWORK TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411246049.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-06
Publication Date
2024-11-15
Estimated Expiration
2044-09-06

AI Technical Summary

Technical Problem

The existing mobile phone security module fails to effectively identify and manage the relevance of sorting codes in data breaches, causing attackers to exploit this vulnerability to repeatedly attack, steal sensitive information, trigger the risk of financial losses and identity theft, and hackers may be lurking in the Deeper sabotage and espionage are carried out in the system.

Method used

By setting the detection window to monitor data leakage events, using deep learning models to analyze the correlation and clustering characteristics between sorted codes, identify high-risk and low-risk sorted codes, divide them into associative sets and non-associative sets, and Adopt stricter security strategies for high-risk sorting codes, optimize their use, and reduce security threats.

Benefits of technology

It improves the protection effect of high-risk sorting codes, enhances the overall security and stability of the system, prevents large-scale data leakage and fraud caused by the relevance of sorting codes, and ensures the security and privacy of user information.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118761069B_ABST
    Figure CN118761069B_ABST
Patent Text Reader

Abstract

The present invention discloses a mobile phone security module management system and method, which relates to the technical field of mobile phone security module management, and includes the following steps: setting a detection window for monitoring and recording the occurrence of data leakage events, and calculating the total number of data leakage events, and judging whether the data leakage events are frequent events based on the total number of data leakage events occurring within the detection window. The present invention monitors data leakage events through the detection window, and uses a deep learning model to analyze the correlation and clustering characteristics of sorting codes, identifies high-risk and low-risk sorting codes, and divides the sorting codes into correlation and non-correlation sets according to the correlation coefficient evaluation results, adopts stricter security policies for high-risk sorting codes, optimizes their use, and reduces security threats, thereby improving the security and stability of the system and ensuring the security and privacy of user information.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of mobile phone security module management, and in particular to a mobile phone security module management system and method. Background Art

[0002] Mobile security module management refers to a technical solution that deploys security software or hardware modules on mobile devices to protect the security of users' personal information and payment data. Its main functions include managing application permissions, real-time monitoring and analyzing security risks on devices, and detecting and isolating malware and suspicious applications. Through these measures, mobile security module management can provide stronger security protection in scenarios such as users using QR code payment to prevent data from being stolen or used for fraud. In the existing QR code payment system, due to insufficient security precautions, mobile security module management can serve as an effective supplement.

[0003] In the security management of mobile phone security modules in the existing technology, multi-level security measures are usually adopted, including data encryption, access control, anomaly detection and real-time monitoring, to deal with security protection after data leakage. Data encryption is to protect sensitive data through encryption algorithms, so that even if the data is leaked, it is difficult to decrypt and use. Access control ensures that only verified users and applications can access sensitive data through authentication and authorization mechanisms. Anomaly detection and real-time monitoring technology are used to identify and prevent abnormal behaviors, such as abnormal access attempts or data transmission, and trigger alarms and take emergency measures in a timely manner.

[0004] The prior art has the following deficiencies:

[0005] When data breaches occur frequently, existing technologies usually do not deeply analyze and manage the sort codes (such as encryption codes) used for encryption, authentication or identification. If the data breach is caused by the high correlation of these sort codes and the mobile phone security module fails to identify this problem, the consequences can be very serious. Attackers can use this security vulnerability to repeatedly attack and steal a large amount of sensitive information, causing users to face the risk of financial losses and identity theft. What's more serious is that hackers may lurk in the system for a long time to carry out deeper sabotage and espionage. Due to the abuse of sort codes, attackers may also forge legitimate identities, conduct unauthorized access and transactions, bypass existing security authentication mechanisms, and cause large-scale fraud.

[0006] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0007] The purpose of the present invention is to provide a mobile phone security module management system and method, which monitors data leakage events by setting a detection window, and uses a deep learning model to analyze the correlation and clustering characteristics between sorting codes, effectively identifying high-risk and low-risk sorting codes, and dividing the sorting codes into correlation sets and non-correlation sets according to the evaluation results of the correlation coefficients, and adopting stricter security policies for high-risk sorting codes to optimize their use and reduce security threats. This not only improves the protection effect on high-risk sorting codes, but also improves the overall security and stability of the system, prevents large-scale data leakage and fraud caused by the correlation of sorting codes, and ensures the security and privacy of user information, so as to solve the problems in the above-mentioned background technology.

[0008] In order to achieve the above object, the present invention provides the following technical solution: a mobile phone security module management method, comprising the following steps:

[0009] A detection window is set to monitor and record the occurrence of data leakage events, and the total number of data leakage events is counted, and whether the data leakage events are frequent events is determined based on the total number of data leakage events that occur within the detection window;

[0010] After determining that the data leakage incident occurs frequently, extract the sorting codes involved in all data leakage incidents within the detection window, aggregate all the extracted sorting codes, and establish an analysis set;

[0011] Use the deep learning model to perform correlation analysis on all the sorting codes in the sorting code analysis set, and based on the analysis results of the deep learning model, conduct a comprehensive evaluation of the correlation of the sorting codes to identify the potential correlation between the sorting codes;

[0012] According to the evaluation results, the sort code analysis set is divided into a correlation set and an irrelevance set. The sort codes in the correlation set show high correlation, while the sort codes in the irrelevance set have no obvious correlation.

[0013] For the sorting codes divided into relevance sets, based on the results of the relevance evaluation in the relevance sets, adjust their selection weight ratios in the security policy, increase restrictions on high-risk sorting codes, and reduce their use frequency in the security mechanism;

[0014] Based on the adjusted sorting code weights, optimize the sorting code usage strategy, strengthen the use of safe sorting codes, and reduce or eliminate high-risk sorting codes.

[0015] Preferably, a frequency threshold is set to determine whether the data leakage event is a frequent event, specifically:

[0016] The total number of data leakage events within the detection window is compared and analyzed with the frequency threshold. If the total number of data leakage events within the detection window is greater than or equal to the set frequency threshold, the data leakage event is determined to be a frequent event and further security measures are triggered. If the total number of data leakage events within the detection window is less than the set frequency threshold, the data leakage event is determined to be within the normal range and no additional security operations are triggered.

[0017] Preferably, the specific steps of extracting the sort codes involved in all data leakage events in the frequently occurring events and aggregating all the extracted sort codes into an analysis set are as follows:

[0018] After confirming that data leakage incidents are frequent events, collect detailed records related to all data leakage incidents within the detection window;

[0019] After collecting detailed records of all data breaches, filter out information related to the sort code;

[0020] After extracting the sort code, perform deduplication and standardization. Deduplication ensures that each sort code appears only once to avoid repeated calculation and analysis. Standardization unifies the format and representation of the sort code to facilitate subsequent analysis and processing.

[0021] Finally, all extracted, deduplicated, and standardized sort codes are aggregated to create a sort code analysis set.

[0022] Preferably, a deep learning model is used to perform correlation analysis on all sorting codes in the sorting code analysis set. The specific process is as follows:

[0023] The mutual dependence between any two sorting codes and the clustering characteristics of the sorting codes in data leakage incidents are extracted from the sorting code analysis set. The mutual dependence between any two sorting codes refers to the frequency and degree of co-occurrence of two sorting codes in data leakage incidents. The clustering characteristics of sorting codes in data leakage incidents refer to the grouping patterns exhibited by sorting codes in multiple data leakage incidents. After analyzing the mutual dependence between any two sorting codes and the clustering characteristics of sorting codes in data leakage incidents, mutual information quantification values ​​and clustering consistency quantification values ​​are generated respectively. The mutual information quantification values ​​and clustering consistency quantification values ​​are input into a pre-trained deep learning model to generate correlation coefficients, and the correlation of the sorting codes is comprehensively evaluated through the correlation coefficients.

[0024] Preferably, the correlation coefficient generated by performing correlation analysis on all the sorting codes in the sorting code analysis set using the deep learning model is compared with a preset correlation coefficient reference threshold, and the sorting code analysis set is divided. The specific division steps are as follows:

[0025] If the correlation coefficient is greater than or equal to the correlation coefficient reference threshold, the sort code analysis set is divided into a correlation set;

[0026] If the correlation coefficient is less than the correlation coefficient reference threshold, the sort code analysis set is divided into a non-correlation set.

[0027] Preferably, in the sorting code analysis set, the specific steps of generating the mutual information quantization value after analyzing the mutual dependence between any two sorting codes are as follows:

[0028] Collect and prepare the sort code information from the data breach incident and construct two sort codes A and sort code B The joint distribution of , which represents the sort code in the data leakage incident A and sort code B The frequency of co-occurrence, when constructing the joint distribution matrix, is defined as: Sort code A and sort code B The probability matrix of co-occurrence in data breach incidents, Sort code A and sort code B The number of co-occurring events, N is the total number of data leakage events, then the calculation expression of the joint probability distribution is: ;

[0029] After constructing the joint distribution, the marginal distribution probability of the two sort codes is calculated, that is, the sort code A and sort code B The probability of each independent occurrence, marginal distribution is used to understand the frequency of occurrence of a single sort code in all data leakage events, and the calculation expression is: , , where Sort code A The probability of independent occurrence, Sort code B The probability of independent occurrence, Sort code A The total number of occurrences, Sort code B Total number of occurrences;

[0030] Use conditional entropy to evaluate the impact of the appearance of one sort code on the appearance of another sort code. The calculation expression of conditional entropy is: , , where Indicates in sort code B If known, the sort code A The conditional entropy of Indicates in sort codeA If known, the sort code B The conditional entropy of a and b Respectively represent the sort code A and sort code B The specific value of

[0031] Calculate the mutual information quantization value to measure the sorting code A and sort code B The amount of information shared between is calculated as: , where Indicates the sort code A and sort code B The mutual information quantification value between .

[0032] Preferably, in the sorting code analysis set, the specific steps of generating the clustering consistency quantization value after analyzing the clustering characteristics of the sorting codes in the data leakage incident are as follows:

[0033] Construct a feature matrix containing all the sorting codes and their occurrence in each data leakage event. The rows of the matrix represent each sorting code, the columns represent each data leakage event, and the element value in the matrix represents the frequency of occurrence of a sorting code in a certain event.

[0034] Clustering the feature matrix using a clustering algorithm to group the sorting codes so that the sorting codes in the same group have similar performance characteristics in data leakage events;

[0035] For each cluster, the distance distribution between its internal sort codes is calculated to evaluate the compactness within the cluster. The distance can be used to represent the difference or similarity between the sort codes. The calculation expression is: , where For two sort codes and The distance measure between For clustering Internal consistency measures, For clustering The number of sort codes in i and j represents the index of the sort code, In the representation k Indicates k clusters;

[0036] Calculate the cluster consistency quantification value, the calculation expression is: , where represents the quantitative value of cluster consistency, , indicating clustering The weight factor of KRepresents the total number of clusters.

[0037] Preferably, for the sorting codes divided into relevance sets, based on the results of the relevance evaluation in the relevance sets, the selection weight ratios of the sorting codes in the security policy are adjusted to increase restrictions on high-risk sorting codes and reduce their use frequency in the security mechanism. The specific steps are as follows:

[0038] According to the sorting codes in the correlation set and the correlation coefficients they generate, the risk weight of each sorting code is calculated. The risk weight reflects the risk level of the sorting code being exploited in a data leakage incident. The calculation expression is: , where Sort code p The risk weight of Sort code p The correlation coefficient represents the comprehensive correlation of the sort code in the correlation set. L is the total number of sort codes in the associativity set;

[0039] According to the risk weights of all sort codes, the overall risk index of the association set is calculated to reflect the security risk level of the entire set. The calculation expression is: , where is the overall risk index of the correlation set;

[0040] According to the calculated risk weight and overall risk index, the selection weight ratio of each sort code in the association set is adjusted. The goal is to reduce the frequency of use of high-risk sort codes and increase restrictions on them. The adjusted expression is: , where Sort code p The selection weight ratio;

[0041] Calculate the risk adjustment factor for each sort code to further adjust the restriction strength of the sort code in the security policy. The calculation expression is: , where Sort code p The risk adjustment factor, is the adjustment coefficient, used to control the adjustment intensity;

[0042] Based on the calculated risk adjustment factor, the final use policy control value is calculated, that is, the use frequency of the sort code in the security policy, and the sort codes in the association set are hierarchically controlled to reduce the use frequency of high-risk sort codes. The calculation expression is: , where Sort code p The final use policy control value, sort code p The final use policy control value Indicates the frequency of use of the sort code in security policies.

[0043] Mobile phone security module management system, including data monitoring module, data extraction and analysis set construction module, deep learning analysis module, sorting code classification module, security policy adjustment module and usage policy optimization module;

[0044] The data monitoring module sets a detection window to monitor and record the occurrence of data leakage events, and calculates the total number of data leakage events, and determines whether the data leakage events are frequent events based on the total number of data leakage events that occur within the detection window;

[0045] The data extraction and analysis set building module, after determining that the data leakage incident is a frequent occurrence, extracts the sorting codes involved in all data leakage incidents within the detection window, aggregates all the extracted sorting codes, and establishes an analysis set;

[0046] The deep learning analysis module uses a deep learning model to perform correlation analysis on all sorting codes in the sorting code analysis set. Based on the analysis results of the deep learning model, the correlation of the sorting codes is comprehensively evaluated to identify the potential correlation between the sorting codes.

[0047] The sorting code classification module divides the sorting code analysis set into a correlation set and a non-correlation set according to the evaluation results. The sorting codes in the correlation set show high correlation, while the sorting codes in the non-correlation set have no obvious correlation.

[0048] The security policy adjustment module adjusts the selection weight ratio of the sorting codes divided into relevance sets in the security policy based on the relevance evaluation results in the relevance sets, increases restrictions on high-risk sorting codes, and reduces their use frequency in the security mechanism;

[0049] Use the strategy optimization module to optimize the sorting code usage strategy based on the adjusted sorting code weights, strengthen the use of safe sorting codes, and reduce or eliminate high-risk sorting codes.

[0050] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0051] The present invention monitors data leakage events by setting a detection window, extracts and analyzes the correlation of sorting codes, and uses a deep learning model to comprehensively evaluate the potential dependencies and clustering characteristics between sorting codes, so as to accurately identify high-risk sorting codes and low-risk sorting codes. According to the evaluation results of the correlation coefficient, the sorting codes are divided into correlation sets and non-correlation sets, and differentiated security strategies are adopted for different sets, restrictions on high-risk sorting codes are increased, and the use strategy of sorting codes is optimized. This can not only improve the protection effect on high-risk sorting codes and reduce potential security threats, but also improve the security and stability of the entire system, avoid large-scale data leakage and fraud caused by the correlation of sorting codes, and by dynamically adjusting security strategies, the system can more flexibly respond to different types of security threats to ensure the security and privacy of user information. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0053] Figure 1 The present invention is a flowchart of the method for managing a mobile phone security module.

[0054] Figure 2 The figure is a module diagram of the mobile phone security module management system of the present invention. DETAILED DESCRIPTION

[0055] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0056] The present invention provides Figure 1 The mobile phone security module management method shown includes the following steps:

[0057] A detection window is set to monitor and record the occurrence of data leakage events, and the total number of data leakage events is counted, and whether the data leakage events are frequent events is determined based on the total number of data leakage events that occur within the detection window;

[0058] The detection window, that is, a specific time period, is used to monitor and record the occurrence of data leakage incidents. Within this window, the system will record all data leakage incidents and calculate the total number of incidents. The setting of the detection window helps to identify the frequency of data leakage and determine whether these incidents are frequent occurrences, so as to determine whether further measures are needed.

[0059] Once the total number of data breach incidents within the detection window is counted, a frequency threshold needs to be set to determine whether the data breach incident is a frequent incident. The frequency threshold is a pre-set benchmark value, usually determined based on historical data analysis, industry standards, or security policies. For example, a system may set a frequency threshold of 10 times per week. If the number of data breach incidents exceeds this frequency threshold, it can be considered that the incident occurs frequently. The key to setting a frequency threshold is to balance security and actual operational needs, to be able to respond to potential threats in a timely manner, and to avoid false alarms caused by oversensitivity.

[0060] After completing the statistics of the total number of data leakage events and setting the frequency threshold, a comparison analysis is performed. If the total number of data leakage events within the detection window is greater than or equal to the set frequency threshold, the system will determine that these events are frequent events and trigger further security measures. If the total number of events is less than the frequency threshold, the data leakage is considered to be within the normal range and no additional security operations are triggered. This comparison process is the core of the entire judgment mechanism. Through simple numerical comparison, potential security threats can be quickly and effectively identified, providing a basis for subsequent security protection.

[0061] After determining that the data leakage incident occurs frequently, extract the sorting codes involved in all data leakage incidents within the detection window, aggregate all the extracted sorting codes, and establish an analysis set;

[0062] Sorting codes are usually used for encryption, authentication or identification, and are an important part of data protection. The purpose of extracting all sorting codes is to further analyze the role of these codes in data leakage incidents, understand whether they have regularity or correlation, and thus discover potential security vulnerabilities. The analysis collection is a centralized management of all sorting codes of data leakage incidents, which is convenient for subsequent correlation analysis. The purpose of this step is to bring together all sorting codes that may be related to data leakage incidents to form a complete analysis sample, in preparation for subsequent correlation analysis using deep learning models.

[0063] After determining that data leakage events occur frequently, extract the sort codes involved in all data leakage events within the detection window, aggregate all extracted sort codes, and establish an analysis set. The specific steps are as follows:

[0064] After confirming that data leakage incidents are frequent events, collect detailed records related to all data leakage incidents within the detection window;

[0065] These records usually include information such as the time of the incident, the type of incident, the scope of data affected, and the sort codes involved. Collecting these detailed records is to ensure that all sort codes related to the data breach can be accurately identified and extracted. The key to this step is the integrity and accuracy of the data to ensure that no potential security threat factors are missed.

[0066] After collecting detailed records of all data breaches, filter out information related to the sort code;

[0067] Sorting codes may be used for data encryption, user authentication, or system identification, so finding all the sorting codes involved in these functions in the records of data breach incidents is a key task. The filtering process usually relies on data analysis tools and log parsing technology to ensure accurate extraction of information related to the sorting code and not miss any key data that may lead to security vulnerabilities.

[0068] Since there may be multiple data leakage events within a detection window, and the same sort code may appear multiple times in different events, after extracting the sort code, it is necessary to perform deduplication and standardization. Deduplication is to ensure that each sort code appears only once to avoid repeated calculation and analysis. Standardization is to unify the format and representation method of the sort code to facilitate subsequent analysis and processing;

[0069] The purpose of this step is to ensure the accuracy and consistency of the analyzed data and reduce the interference of noisy data.

[0070] Finally, all extracted, deduplicated, and standardized ranking codes are aggregated to create a ranking code analysis set;

[0071] This set contains all the sort codes related to data leakage incidents within the detection window, providing a complete data foundation for subsequent correlation analysis. The purpose of establishing a sort code analysis set is to centrally manage and analyze these codes, identify potential security vulnerabilities and risk factors, and ensure that the optimization of security policies can be based on evidence. By analyzing the set, the correlation between sort codes can be effectively identified, and security policies can be adjusted accordingly.

[0072] Use the deep learning model to perform correlation analysis on all the sorting codes in the sorting code analysis set, and based on the analysis results of the deep learning model, conduct a comprehensive evaluation of the correlation of the sorting codes to identify the potential correlation between the sorting codes;

[0073] In this way, it is possible to identify which sorting codes are more likely to be exploited in data breaches, find security vulnerabilities, and provide a basis for further security policy adjustments.

[0074] Use the deep learning model to perform correlation analysis on all the sort codes in the sort code analysis set. The specific process is as follows:

[0075] The mutual dependence between any two sorting codes and the clustering characteristics of the sorting codes in data leakage incidents are extracted from the sorting code analysis set. The mutual dependence between any two sorting codes refers to the frequency and degree of co-occurrence of two sorting codes in data leakage incidents, reflecting whether there is a strong dependence relationship between the two. The clustering characteristics of the sorting codes in data leakage incidents refer to the grouping pattern shown by the sorting codes in multiple data leakage incidents, reflecting which sorting codes are more likely to co-appear in the same type of events. After analyzing the mutual dependence between any two sorting codes and the clustering characteristics of the sorting codes in data leakage incidents, mutual information quantification values ​​and clustering consistency quantification values ​​are generated respectively, and the mutual information quantification values ​​and clustering consistency quantification values ​​are input into the pre-trained deep learning model to generate correlation coefficients, and the correlation of the sorting codes is comprehensively evaluated through the correlation coefficients.

[0076] A pre-trained deep learning model is a neural network model that has been learned and optimized through a large amount of historical data, and can automatically identify and extract complex data patterns and features. The deep learning model has been trained to analyze the interdependence and clustering characteristics between sort codes. By inputting the mutual information quantification value and the cluster consistency quantification value, the model can generate a correlation coefficient to quantify the strength of the correlation between sort codes. This model usually contains a multi-layer neuron structure that can learn nonlinear relationships and high-dimensional features, enabling it to accurately evaluate the correlation between sort codes.

[0077] If there is a strong interdependence between any two sort codes in the sort code analysis set, this usually indicates that there is a strong correlation between the sort codes. Strong interdependence means that the two sort codes frequently appear together in data leakage incidents. This co-occurrence pattern suggests that they may be used together or have similar effects in certain attacks or data leakage incidents. The high frequency of interdependence between multiple sort codes indicates that they are associated and exploited under specific conditions, and may form a regular combination pattern that is exploited by attackers. Therefore, this interdependence not only reflects the direct correlation between the sort codes in specific events, but also implies their overall correlation in the entire analysis set. High correlation may mean that there is a certain logical or functional connection between the sort codes, which makes them show high consistency and synergy in data leakage incidents. In this case, attackers may use this correlation to carry out more precise attacks, increasing the risk and destructive power of data leakage.

[0078] In the sorting code analysis set, the specific steps of generating the mutual information quantization value after analyzing the mutual dependence between any two sorting codes are as follows:

[0079] Collect and prepare the sort code information from the data breach incident and construct two sort codes A and sort code B The joint distribution of , which represents the sort code in the data leakage incident A and sort code B The frequency of co-occurrence, when constructing the joint distribution matrix, is defined as: Sort code A and sort code B The probability matrix of co-occurrence in data breach incidents, Sort code A and sort code B The number of co-occurring events, N is the total number of data leakage events, then the calculation expression of the joint probability distribution is: ;

[0080] After constructing the joint distribution, the marginal distribution probability of the two sort codes is calculated, that is, the sort code A The probability of A and sort code B appearing independently. The marginal distribution is used to understand the frequency of occurrence of a single sort code in all data leakage events. The calculation expression is: , , where Sort code A The probability of independent occurrence, Sort code B The probability of independent occurrence, Sort code A The total number of occurrences, Sort code B Total number of occurrences;

[0081] Conditional entropy measures the uncertainty of the occurrence of another sorting code when the occurrence of one sorting code is known. Use conditional entropy to evaluate the impact of the occurrence of one sorting code on the occurrence of another sorting code. The calculation expression of conditional entropy is: , , where Indicates in sort code B If known, the sort code A The conditional entropy of Indicates in sort code A If known, the sort code B The conditional entropy of a and b Respectively represent the sort code A and sort code B The specific value of Indicates the sort code A You can get one value among all specific values a This means a Is the sort code A A specific instance or possible state of a.

[0082] Indicates the sort code B You can get one value among all specific values b .same, b Is the sort code B A specific instance or possible state of a.

[0083] Calculate the mutual information quantization value to measure the sorting code A and sort code B The amount of information shared between is calculated as: , where Indicates the sort code A and sort code B The mutual information quantification value between .

[0084] The mutual information quantization value can be used to determine the strength of the correlation between the sorting codes. The mutual information quantization value quantifies the mutual dependence between the two sorting codes, reflecting the degree to which the appearance of one sorting code reduces the information uncertainty of the other sorting code. Therefore, if the mutual information quantization value generated after analyzing any two sorting codes is larger, it means that the two sorting codes share more information and their correlation is stronger, indicating that the two sorting codes are likely to be used together or have similar behavioral characteristics in the data leakage incident. Conversely, if the mutual information quantization value has a smaller performance value, it means that the amount of information shared between the sorting codes is less, their correlation is not strong, and they may appear alone in the data leakage incident or show independent characteristics. This analysis helps to identify and understand the relationship between the sorting codes, thereby optimizing security protection measures and reducing the risk of data leakage.

[0085] In the analysis set of sort codes, if the sort codes show high consistency in the cluster analysis, that is, these sort codes often appear in the same group, this indicates that the correlation between the sort codes is strong. This high consistency means that these sort codes frequently co-appear in multiple data leakage incidents, showing a certain pattern or regularity. In other words, there is a potential relationship between the sort codes, perhaps because they are repeatedly used by the same attackers, or they have similar susceptibility to the same type of attacks. This situation usually indicates that the sort code combination may constitute a specific security vulnerability or a target preferred by the attacker, so they have a higher correlation. The clustering characteristics with high consistency reflect the consistency and co-occurrence of sort codes in different data leakage incidents. This consistency may be due to the logical relationship between the sort codes, the similarity of usage scenarios, or the same weaknesses exploited. By analyzing these clustering characteristics, the potential dependencies and risks between the sort codes can be identified, which will help to further optimize security protection measures and prevent the recurrence of data leakage incidents.

[0086] In the sort code analysis set, the specific steps for analyzing the clustering characteristics of the sort codes in the data leakage incidents and generating the cluster consistency quantification value are as follows:

[0087] Construct a feature matrix containing all sort codes and their occurrence in each data leakage event. The rows of the matrix represent each sort code, the columns represent each data leakage event, and the element value in the matrix represents the frequency of occurrence of a sort code in a certain event.

[0088] Use clustering algorithms (such as k-means clustering, hierarchical clustering, etc.) to cluster the feature matrix and group the sort codes so that the sort codes in the same group have similar performance characteristics in data leakage incidents;

[0089] Making the performance characteristics of the ranking codes in the same group similar in data breach incidents means assigning the ranking codes with similar patterns or behaviors to the same group through clustering algorithms (such as k-means clustering or hierarchical clustering). Similar performance characteristics mean that these ranking codes appear or act in similar ways in different data breach incidents. For example, these ranking codes may frequently appear at the same time in the same incident, or their frequency of occurrence has a similar distribution in multiple incidents. By grouping these ranking codes into the same cluster, we can identify groups of ranking codes with similar attack patterns or exploited by attackers in similar ways, so as to more effectively analyze and respond to security threats.

[0090] For each cluster, the distance distribution between its internal sort codes is calculated to evaluate the compactness within the cluster. The distance can be used to represent the difference or similarity between the sort codes. The calculation expression is: , where For two sort codes and The distance metric between them can be based on cosine similarity or other non-standard deviation distance formulas. For clustering Internal consistency measures, For clustering The number of sort codes in i and j represents the index of the sort code, In the representation k Indicates k clusters;

[0091] Calculate the cluster consistency quantification value, the calculation expression is: , where represents the quantitative value of cluster consistency, , indicating clustering The weight factor relative to the total number of sort codes n In terms of K Represents the total number of clusters.

[0092] It can be seen from the cluster consistency quantification value that the larger the performance value of the cluster consistency quantification value, the stronger the correlation between the sorting codes in the sorting code analysis set. The cluster consistency quantification value measures the stability and compactness of the clusters formed by the sorting codes in data leakage incidents. When the cluster consistency quantification value is high, it means that the sorting codes show a high degree of grouping consistency in multiple data leakage incidents, reflecting that there are strong mutual correlations and common characteristic patterns between the sorting codes. On the contrary, if the cluster consistency quantification value is low, it means that the distribution of sorting codes in data leakage incidents is more dispersed and inconsistent, and the correlation between sorting codes is weaker. This index helps identify the clustering behavior of sorting codes and reveals their potential connections and synergies in data leakage incidents.

[0093] The deep learning model is not specifically limited here, and can realize the quantization of mutual information and cluster consistency quantification value Perform comprehensive analysis to generate correlation coefficients In order to realize the technical solution of the present invention, the present invention provides a specific implementation method;

[0094] Correlation coefficient The resulting calculation formula is: , where , The mutual information quantization values ​​are and cluster consistency quantification value The preset scaling factor of , Both are greater than 0.

[0095] It can be seen from the correlation coefficient that the greater the performance value of the mutual information quantization value generated after analyzing the mutual dependence between any two sorting codes, the greater the performance value of the clustering consistency quantization value generated after analyzing the clustering characteristics of the sorting codes in the data leakage incident. That is, the greater the performance value of the correlation coefficient generated by using the deep learning model to perform correlation analysis on all the sorting codes in the sorting code analysis set, the stronger the correlation between the sorting codes in the sorting code analysis set, and vice versa.

[0096] According to the evaluation results, the sort code analysis set is divided into a correlation set and an irrelevance set. The sort codes in the correlation set show high correlation, while the sort codes in the irrelevance set have no obvious correlation.

[0097] This division helps to adopt different security strategies for different sort codes and improve the protection effect.

[0098] The correlation coefficient generated by using the deep learning model to perform correlation analysis on all the sorting codes in the sorting code analysis set is compared with the pre-set correlation coefficient reference threshold, and the sorting code analysis set is divided. The specific division steps are as follows:

[0099] If the correlation coefficient is greater than or equal to the correlation coefficient reference threshold, the sort code analysis set is divided into a correlation set;

[0100] If the correlation coefficient is less than the correlation coefficient reference threshold, the sort code analysis set is divided into a non-correlation set.

[0101] When the sort code analysis set is divided into a non-correlated set, it means that the sort codes involved in these data leakage incidents do not show significant correlation or dependence. This means that the appearance of these sort codes in the data leakage incident is independent, there is no specific pattern or regularity between each other, and there are no common characteristics or uses. Therefore, the data leakage incident may be caused by a variety of independent and unrelated factors, rather than because of the combination or association between certain sort codes. Such analysis results show that there are currently no security vulnerabilities with systematic or synergistic effects between sort codes, and it may be necessary to focus on security protection strategies in other dimensions.

[0102] For the sorting codes divided into relevance sets, based on the results of the relevance evaluation in the relevance sets, adjust their selection weight ratios in the security policy, increase restrictions on high-risk sorting codes, and reduce their use frequency in the security mechanism;

[0103] The purpose of this step is to reduce the probability of high-risk sort codes being abused, thereby improving the overall security of the system.

[0104] For the sorting codes divided into relevance sets, based on the results of the relevance evaluation in the relevance sets, adjust their selection weight ratios in the security policy, increase restrictions on high-risk sorting codes, and reduce their use frequency in the security mechanism. The specific steps are as follows:

[0105] According to the sorting codes in the correlation set and the correlation coefficients they generate, the risk weight of each sorting code is calculated. The risk weight reflects the risk level of the sorting code being exploited in a data leakage incident. The calculation expression is: , where Sort code p The risk weight of Sort code p The correlation coefficient represents the comprehensive correlation of the sort code in the correlation set. L is the total number of sort codes in the associativity set;

[0106] Risk Weight By sort code p It is calculated by dividing the square of the correlation coefficient of the sort code by the sum of the square of the correlation coefficients of all sort codes. This method emphasizes the sort codes with high correlation coefficients and increases their proportion in the total risk;

[0107] Sort code The correlation coefficient reflects the comprehensive correlation of the sort code in the correlation set, which can be calculated by comprehensively analyzing the interdependence and clustering characteristics of the sort code in different data leakage events. Specifically, this correlation coefficient needs to take into account the sort code p The frequency of co-occurrence with other ranking codes in data breach incidents and the grouping patterns and consistency of these ranking codes in cluster analysis. By inputting mutual information quantification values ​​and cluster consistency quantification values ​​into the deep learning model, the model automatically learns the nonlinear combination and mutual influence of these features, thereby generating a correlation coefficient that reflects the overall correlation of the ranking codes. , which will not be described in detail here.

[0108] According to the risk weights of all sort codes, the overall risk index of the association set is calculated to reflect the security risk level of the entire set. The calculation expression is: , where is the overall risk index of the correlation set;

[0109] Here, the overall risk index The risk weights are calculated using a logarithmic form of a weighted exponential function, emphasizing the nonlinear contribution of high-risk ranking codes to the overall risk.

[0110] According to the calculated risk weight and overall risk index, the selection weight ratio of each sort code in the association set is adjusted. The goal is to reduce the frequency of use of high-risk sort codes and increase restrictions on them. The adjusted expression is: , where Sort code p The selection weight ratio;

[0111] Select weight ratio Use logical functions to adjust risk weights to ensure that high-risk sorting codes have lower weight ratios and limit their use in security policies.

[0112] Calculate the risk adjustment factor for each sort code to further adjust the restriction strength of the sort code in the security policy. The calculation expression is: , where Sort code p The risk adjustment factor, It is the adjustment coefficient, which is used to control the adjustment intensity. It is not specifically limited here and can be adjusted according to the actual situation.

[0113] Risk Adjustment Factors By choosing the weight ratio Multiply by the adjustment factor The sinusoidal function calculation of the correlation coefficient allows nonlinear adjustment of the usage restrictions of the sorting code to adapt to different security requirements.

[0114] Based on the calculated risk adjustment factor, the final use policy control value is calculated, that is, the use frequency of the sort code in the security policy, and the sort codes in the association set are hierarchically controlled to reduce the use frequency of high-risk sort codes. The calculation expression is: , where Sort code p The final use policy control value, sort code p The final use policy control value Indicates the frequency of use of the sort code in the security policy;

[0115] Final use policy control value The risk adjustment factor of the sort code is calculated by dividing the risk adjustment factor of all sort codes, ensuring that the use of sort codes in security policies is properly controlled according to their risk adjustment factors. This method can effectively prevent data leakage risks by comprehensively evaluating the risks of sort codes and dynamically adjusting security policies.

[0116] Based on the adjusted sort code weights, optimize the sort code usage strategy, strengthen the use of safe sort codes, and reduce or eliminate high-risk sort codes;

[0117] After evaluating the relevance and risk level of the sorting codes, the weight of each sorting code is adjusted to optimize its usage strategy in the system. Specifically, the adjusted weights reflect the security and risk of the sorting codes. Based on these weights, the system will give priority to sorting codes that are assessed to be more secure, while reducing or eliminating the frequency of use of sorting codes that are considered to be high-risk. This optimization strategy aims to improve the overall security of the system, avoid potential security vulnerabilities and data leakage risks, and ensure that the system processes and manages data more securely during operation.

[0118] The new sort code usage strategy should take into account the risk characteristics of the association set and enhance the effectiveness of data protection by selecting more secure sort codes to prevent the recurrence of data leakage incidents.

[0119] The present invention monitors data leakage events by setting a detection window, extracts and analyzes the correlation of sorting codes, and uses a deep learning model to comprehensively evaluate the potential dependencies and clustering characteristics between sorting codes, so as to accurately identify high-risk sorting codes and low-risk sorting codes. According to the evaluation results of the correlation coefficient, the sorting codes are divided into correlation sets and non-correlation sets, and differentiated security strategies are adopted for different sets, restrictions on high-risk sorting codes are increased, and the use strategy of sorting codes is optimized. This can not only improve the protection effect on high-risk sorting codes and reduce potential security threats, but also improve the security and stability of the entire system, avoid large-scale data leakage and fraud caused by the correlation of sorting codes, and by dynamically adjusting security strategies, the system can more flexibly respond to different types of security threats to ensure the security and privacy of user information.

[0120] The present invention provides Figure 2 The mobile phone security module management system shown includes a data monitoring module, a data extraction and analysis set construction module, a deep learning analysis module, a sorting code classification module, a security policy adjustment module, and a usage policy optimization module;

[0121] The data monitoring module sets a detection window to monitor and record the occurrence of data leakage events, and calculates the total number of data leakage events, and determines whether the data leakage events are frequent events based on the total number of data leakage events that occur within the detection window;

[0122] The data extraction and analysis set building module, after determining that the data leakage incident is a frequent occurrence, extracts the sorting codes involved in all data leakage incidents within the detection window, aggregates all the extracted sorting codes, and establishes an analysis set;

[0123] The deep learning analysis module uses a deep learning model to perform correlation analysis on all sorting codes in the sorting code analysis set. Based on the analysis results of the deep learning model, the correlation of the sorting codes is comprehensively evaluated to identify the potential correlation between the sorting codes.

[0124] The sorting code classification module divides the sorting code analysis set into a correlation set and a non-correlation set according to the evaluation results. The sorting codes in the correlation set show high correlation, while the sorting codes in the non-correlation set have no obvious correlation.

[0125] The security policy adjustment module adjusts the selection weight ratio of the sorting codes divided into relevance sets in the security policy based on the relevance evaluation results in the relevance sets, increases restrictions on high-risk sorting codes, and reduces their use frequency in the security mechanism;

[0126] Use the strategy optimization module to optimize the use strategy of sorting codes based on the adjusted sorting code weights, strengthen the use of safe sorting codes, and reduce or eliminate high-risk sorting codes;

[0127] The mobile phone security module management method provided in the embodiment of the present invention is implemented by the mobile phone security module management system mentioned above. The specific methods and processes of the mobile phone security module management system are detailed in the embodiment of the mobile phone security module management method mentioned above, which will not be repeated here.

[0128] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0129] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0130] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A mobile phone security module management method, characterized in that: The following steps are involved: A detection window is set to monitor and record the occurrence of data leakage events, and the total number of data leakage events is counted, and whether the data leakage events are frequent events is determined based on the total number of data leakage events that occur within the detection window; After determining that the data leakage incident occurs frequently, extract the sorting codes involved in all data leakage incidents within the detection window, aggregate all the extracted sorting codes, and establish an analysis set; Use the deep learning model to perform correlation analysis on all the sorting codes in the sorting code analysis set, and based on the analysis results of the deep learning model, conduct a comprehensive evaluation of the correlation of the sorting codes to identify the potential correlation between the sorting codes; According to the evaluation results, the sort code analysis set is divided into a correlation set and an irrelevance set. The sort codes in the correlation set show high correlation, while the sort codes in the irrelevance set have no obvious correlation. For the sorting codes divided into relevance sets, based on the results of the relevance evaluation in the relevance sets, adjust their selection weight ratios in the security policy, increase restrictions on high-risk sorting codes, and reduce their use frequency in the security mechanism; Based on the adjusted sorting code weights, optimize the sorting code usage strategy, strengthen the use of safe sorting codes, and reduce or eliminate high-risk sorting codes.

2. The mobile phone security module management method according to claim 1, characterized in that: Set a frequency threshold to determine whether a data breach is a frequent event, specifically: The total number of data leakage events within the detection window is compared and analyzed with the frequency threshold. If the total number of data leakage events within the detection window is greater than or equal to the set frequency threshold, the data leakage event is determined to be a frequent event and further security measures are triggered. If the total number of data leakage events within the detection window is less than the set frequency threshold, the data leakage event is determined to be within the normal range and no additional security operations are triggered.

3. The mobile phone security module management method according to claim 1, characterized in that: The specific steps for extracting the sort codes involved in all data leakage events in the frequently occurring events and aggregating all the extracted sort codes into an analysis set are as follows: After confirming that data leakage incidents are frequent events, collect detailed records related to all data leakage incidents within the detection window; After collecting detailed records of all data breaches, filter out information related to the sort code; After extracting the sort code, perform deduplication and standardization. Deduplication ensures that each sort code appears only once to avoid repeated calculation and analysis. Standardization unifies the format and representation of the sort code to facilitate subsequent analysis and processing. Finally, all extracted, deduplicated, and standardized sort codes are aggregated to create a sort code analysis set.

4. The mobile phone security module management method according to claim 1, characterized in that: Use the deep learning model to perform correlation analysis on all the sort codes in the sort code analysis set. The specific process is as follows: The mutual dependence between any two sorting codes and the clustering characteristics of the sorting codes in data leakage incidents are extracted from the sorting code analysis set. The mutual dependence between any two sorting codes refers to the frequency and degree of co-occurrence of two sorting codes in data leakage incidents. The clustering characteristics of sorting codes in data leakage incidents refer to the grouping patterns exhibited by sorting codes in multiple data leakage incidents. After analyzing the mutual dependence between any two sorting codes and the clustering characteristics of sorting codes in data leakage incidents, mutual information quantification values ​​and clustering consistency quantification values ​​are generated respectively. The mutual information quantification values ​​and clustering consistency quantification values ​​are input into a pre-trained deep learning model to generate correlation coefficients, and the correlation of the sorting codes is comprehensively evaluated through the correlation coefficients.

5. The method for managing a mobile phone security module according to claim 4, characterized in that: The correlation coefficient generated by using the deep learning model to perform correlation analysis on all the sorting codes in the sorting code analysis set is compared with the pre-set correlation coefficient reference threshold, and the sorting code analysis set is divided. The specific division steps are as follows: If the correlation coefficient is greater than or equal to the correlation coefficient reference threshold, the sort code analysis set is divided into a correlation set; If the correlation coefficient is less than the correlation coefficient reference threshold, the sort code analysis set is divided into a non-correlation set.

6. The method for managing a mobile phone security module according to claim 4, characterized in that: In the sorting code analysis set, the specific steps of generating the mutual information quantization value after analyzing the mutual dependence between any two sorting codes are as follows: Collect and prepare the sort code information from the data breach incident and construct two sort codes A and sort code B The joint distribution of , which represents the sort code in the data leakage incident A and sort code B The frequency of co-occurrence, when constructing the joint distribution matrix, is defined as: Sort code A and sort code B The probability matrix of co-occurrence in data breach incidents, Sort code A and sort code B The number of co-occurring events, N is the total number of data leakage events, then the calculation expression of the joint probability distribution is: ; After constructing the joint distribution, the marginal distribution probability of the two sort codes is calculated, that is, the sort code A and sort code B The probability of each independent occurrence, marginal distribution is used to understand the frequency of occurrence of a single sort code in all data leakage events, and the calculation expression is: , , where Sort code A The probability of independent occurrence, Sort code B The probability of independent occurrence, Sort code A The total number of occurrences, Sort code B Total number of occurrences; Conditional entropy is used to evaluate the impact of the appearance of one sort code on the appearance of another sort code. The calculation expression of conditional entropy is: , , where Indicates in sort code B If known, the sort code A The conditional entropy of Indicates in sort code A If known, the sort code B The conditional entropy of a and b Respectively represent the sort code A and sort code B The specific value of Calculate the mutual information quantization value to measure the sorting code A and sort code B The amount of information shared between is calculated as: , where Indicates the sort code A and sort code B The mutual information quantification value between .

7. The method for managing a mobile phone security module according to claim 4, characterized in that: In the sort code analysis set, the specific steps for generating cluster consistency quantification values ​​after analyzing the clustering characteristics of the sort codes in the data leakage incident are as follows: Construct a feature matrix containing all the sorting codes and their occurrence in each data leakage event. The rows of the matrix represent each sorting code, the columns represent each data leakage event, and the element value in the matrix represents the frequency of occurrence of a sorting code in a certain event. Clustering the feature matrix using a clustering algorithm to group the sorting codes so that the sorting codes in the same group have similar performance characteristics in data leakage events; For each cluster, the distance distribution between its internal sorting codes is calculated to evaluate the compactness within the cluster. The distance can be used to represent the difference or similarity between the sorting codes. The calculation expression is: , where For two sort codes and The distance measure between For clustering Internal consistency measures, For clustering The number of sort codes in i and j represents the index of the sort code, In the representation k Indicates k clusters; Calculate the cluster consistency quantification value, the calculation expression is: , where represents the quantitative value of cluster consistency, , indicating clustering The weight factor of K Represents the total number of clusters.

8. The method for managing a mobile phone security module according to claim 4, characterized in that: For the sorting codes divided into relevance sets, based on the results of the relevance evaluation in the relevance sets, adjust their selection weight ratios in the security policy, increase restrictions on high-risk sorting codes, and reduce their use frequency in the security mechanism. The specific steps are as follows: According to the sorting codes in the correlation set and the correlation coefficients they generate, the risk weight of each sorting code is calculated. The risk weight reflects the risk level of the sorting code being exploited in a data leakage incident. The calculation expression is: , where Sort code p The risk weight of Sort code p The correlation coefficient represents the comprehensive correlation of the sort code in the correlation set. L is the total number of sort codes in the associativity set; According to the risk weights of all sort codes, the overall risk index of the association set is calculated to reflect the security risk level of the entire set. The calculation expression is: , where is the overall risk index of the correlation set; According to the calculated risk weight and overall risk index, the selection weight ratio of each sort code in the association set is adjusted. The goal is to reduce the frequency of use of high-risk sort codes and increase restrictions on them. The adjusted expression is: , where Sort code p The selection weight ratio; Calculate the risk adjustment factor for each sort code to further adjust the restriction strength of the sort code in the security policy. The calculation expression is: , where Sort code p The risk adjustment factor, is the adjustment coefficient, used to control the adjustment intensity; Based on the calculated risk adjustment factor, the final use policy control value is calculated, that is, the use frequency of the sort code in the security policy, and the sort codes in the association set are hierarchically controlled to reduce the use frequency of high-risk sort codes. The calculation expression is: , where Sort code p The final use policy control value, sort code p The final use policy control value Indicates the frequency of use of the sort code in security policies.

9. A mobile phone security module management system, used to implement the mobile phone security module management method according to any one of claims 1 to 8, characterized in that: It includes data monitoring module, data extraction and analysis set building module, deep learning analysis module, sorting code classification module, security policy adjustment module and usage policy optimization module; The data monitoring module sets a detection window to monitor and record the occurrence of data leakage events, and calculates the total number of data leakage events, and determines whether the data leakage events are frequent events based on the total number of data leakage events that occur within the detection window; The data extraction and analysis set building module, after determining that the data leakage incident is a frequent occurrence, extracts the sorting codes involved in all data leakage incidents within the detection window, aggregates all the extracted sorting codes, and establishes an analysis set; The deep learning analysis module uses a deep learning model to perform correlation analysis on all sorting codes in the sorting code analysis set. Based on the analysis results of the deep learning model, the correlation of the sorting codes is comprehensively evaluated to identify the potential correlation between the sorting codes. The sorting code classification module divides the sorting code analysis set into a correlation set and a non-correlation set according to the evaluation results. The sorting codes in the correlation set show high correlation, while the sorting codes in the non-correlation set have no obvious correlation. The security policy adjustment module adjusts the selection weight ratio of the sorting codes divided into relevance sets in the security policy based on the relevance evaluation results in the relevance sets, increases restrictions on high-risk sorting codes, and reduces their use frequency in the security mechanism; Use the strategy optimization module to optimize the sorting code usage strategy based on the adjusted sorting code weights, strengthen the use of safe sorting codes, and reduce or eliminate high-risk sorting codes.