Data access permission dynamic adjustment system

Abstract

The invention discloses a data access permission dynamic adjustment system, which belongs to the technical field of data management and comprises a data collection module, a feature extraction module, a risk assessment module, a strategy generation module, a decision optimization module, a verification adjustment module, a permission application module, a user feedback module, an anomaly detection module and a monitoring recording module. According to the method, the candidate permission strategies can be quickly generated, the real-time decision-making requirement can be met, meanwhile, more selection spaces are provided for subsequent selection, a user is helped to understand the basis for strategy generation more visually, the permission management strategies can be optimized and adjusted easily, the access path can be adjusted in real time, the flexibility of the system is improved, and the user experience is improved. Global search can be carried out in a complex permission space, local optimum is avoided, and access continuity and stability are guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of data management, and in particular to a system for dynamically adjusting data access rights. Background Art

[0002] In the digital age, data security and access control have become important components for protecting sensitive information. With the popularization of cloud computing and big data technologies, enterprises and institutions are facing increasingly complex access management challenges. The effectiveness of data access rights management directly affects the security of information systems and the trust of users. With the continuous changes in organizational structure and technical environment, traditional permission management methods have gradually exposed their shortcomings. Traditional static permission management models often cannot adapt to dynamically changing access requirements in real time, leading to potential security risks and data leakage. In order to solve this problem, it is particularly necessary to develop a data access rights dynamic adjustment system.

[0003] The existing data access permission dynamic adjustment system takes a long time to generate candidate permission policies, which cannot meet the needs of real-time decision-making, and cannot help users understand the basis of policy generation more intuitively, which is not conducive to optimizing and adjusting permission management policies. In addition, the existing data access permission dynamic adjustment system has low flexibility and cannot perform global search in complex permission spaces. It is easy to fall into local optimality, which reduces the continuity and stability of access. To this end, we propose a data access permission dynamic adjustment system. Summary of the invention

[0004] The purpose of the present invention is to solve the defects in the prior art and to propose a system for dynamically adjusting data access rights.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A data access rights dynamic adjustment system, comprising a data collection module, a feature extraction module, a risk assessment module, a strategy generation module, a decision optimization module, a verification adjustment module, a rights application module, a user feedback module, an anomaly detection module and a monitoring and recording module;

[0007] The data collection module is used to collect user behavior data, access requests and network environment information in real time;

[0008] The feature extraction module is used to pre-process the raw data collected by the data collection module and perform feature extraction and dimensionality reduction processing;

[0009] The risk assessment module is used to perform risk assessment on the user's current access request;

[0010] The policy generation module is used to simulate and evaluate different policies based on the risk assessment results to generate a preliminary authority policy;

[0011] The decision optimization module is used to optimize the preliminary permission strategy and dynamically optimize the access path;

[0012] The verification and adjustment module is used to simulate real access requests, verify and adjust policies;

[0013] The permission application module is used to deploy the verified permission policy into the actual data access control system;

[0014] The user feedback module is used to collect users' opinions and suggestions on the current permission policy in real time and optimize the permission adjustment policy;

[0015] The anomaly detection module is used to compare the user's historical behavior and current behavior pattern, and detect possible abnormal access behavior in real time;

[0016] The monitoring and recording module is used to record permission adjustment information and access behavior.

[0017] As a further solution of the present invention, the feature extraction module extracts the dimensionality reduction in the following specific steps:

[0018] S1.1: Remove missing values, duplicate values, and abnormal data from each group of raw data, normalize data of different dimensions to the same range through the standard score normalization method, and input each group of processed data into the pre-trained CNN network;

[0019] S1.2: After receiving each set of input data, the CNN network uses different sizes of convolution kernels in the first convolution layer to perform convolution operations on the input data to extract local features, and capture different aspects of the data based on multiple convolution kernels to generate multiple feature maps. The feature maps output by the convolution layer are nonlinearly activated through the ReLU activation function, and each feature map is element-by-element operated to generate a nonlinear feature map;

[0020] S1.3: The generated nonlinear feature map is passed through the pooling layer for maximum pooling operation, and the maximum value of each area in the feature map is retained to obtain the downsampled feature map. The output of the pooling layer is then input into the new convolution layer for convolution operation again. The convolution, activation and pooling processes are repeated until the preset processing cycle is reached. The final feature map is input, and then the CNN network fully connected layer flattens the multi-dimensional feature map after convolution and pooling into a one-dimensional vector.

[0021] S1.4: The output layer receives the processed one-dimensional vector and post-processes it to output the corresponding groups of feature data. Then, each group of extracted data is projected into a subspace with lower dimensionality through linear transformation, and each group of feature data is normalized through L2 regularization method.

[0022] As a further solution of the present invention, the risk assessment module performs risk assessment on the user's current access request in the following specific steps:

[0023] S2.1: Extract the user's current access request features from each group of feature data after dimensionality reduction, integrate the extracted feature data into a request feature set, collect different risk categories, and calculate the initial estimate of each group of risk categories when no request features are received based on the number of requests belonging to the risk level in the historical data;

[0024] S2.2: The probability distribution of each feature data in the requested feature set under different risk categories is modeled through a multinomial distribution model, and the probability of observing the current feature under each risk category is obtained based on the modeling results. Then, the posterior probability of each risk category being established after observing any feature is calculated based on the Bayesian theorem;

[0025] S2.3: Collect the calculated posterior probabilities of each group and select the category with the highest risk as the current risk assessment result. Dynamically set the thresholds of each risk level based on the actual situation and historical data. Then, adopt the corresponding strategy based on the risk level corresponding to the risk category.

[0026] As a further solution of the present invention, the specific steps of the policy generation module generating a preliminary authority policy are as follows:

[0027] S3.1: Based on the extracted feature data, different system status information is obtained. Based on the risk assessment results, the system can take different sets of actions in each state, where each action corresponds to an authority policy adjustment. The current system state is then used as the root node, and the new states generated after taking different actions in the current state are used as child nodes to construct the corresponding exploration tree.

[0028] S3.2: Calculate the confidence bound value of each node in the exploration tree, and then based on the UCB selection strategy, gradually select the node with the largest confidence bound value until the selected child node is a leaf node that has not been fully visited. Generate the corresponding new state according to the various groups of actions that can be taken under the system state of the current leaf node, and add the generated new state as a new child node to the exploration tree;

[0029] S3.3: Evaluate the benefits of each group of expanded child nodes through random simulation until the terminal state or the preset number of steps is reached, then stop the simulation. After the simulation is completed, trace the simulation results back to all the nodes passed on the path, and update the number of visits and cumulative benefits of each node;

[0030] S3.4: Repeat selection, expansion, simulation, and backtracking until the preset iteration time is reached, collect all node information in the exploration tree, and select the action setting corresponding to the node path with the most visits as the preliminary permission strategy.

[0031] As a further solution of the present invention, the specific steps of optimizing the preliminary authority strategy by the decision optimization module are as follows:

[0032] S4.1: Extract multiple groups of permission policy information from historical data to construct a corresponding policy space, and add the generated preliminary permission policies to the policy space. Then, initialize a population in the policy space by random generation, and each individual in the population represents a group of potential permission policies.

[0033] S4.2: Calculate the fitness value of each individual based on the security, user experience and system performance indicators of the strategy, select the three individuals with the highest fitness values ​​based on the fitness values, and then update the positions of the remaining individuals in the strategy space by simulating hunting behavior;

[0034] S4.3: After updating the individual position, check whether the individual position is out of bounds. If the new position of the individual exceeds the set authority policy range, the individual is corrected, and the fitness value selection, position update and correction are repeated until the fitness converges to the preset range and the iteration is stopped;

[0035] S4.4: After the iteration stops, the individual with the highest fitness is used as the optimized permission strategy, and the final permission strategy is passed to the permission application module, and the user access request is managed according to the final permission strategy.

[0036] As a further solution of the present invention, the decision optimization module dynamically optimizes the access path in the following specific steps:

[0037] S5.1: Construct all resources and access rights in the system into a graph, where nodes represent resources and edges represent feasible access paths. The cost of the access path is used as the weight of each edge, and the pheromone concentration is initialized on each edge. At the same time, the heuristic value of each group of access paths is calculated based on security and response time.

[0038] S5.2: Initialize multiple groups of adjustment individuals, and randomly distribute the positions of the initial groups of adjustment individuals at the starting nodes in the graph, calculate the selection probability of each group of paths in the graph according to the pheromone concentration and the heuristic value, and randomly select the next access path from the starting node according to the selection probability of each path to visit the target node in sequence until the end point is reached;

[0039] S5.3: After the path traversal is completed, the weight pheromones of the edges passed are updated, and the path selection, path construction and pheromone update are repeated until the maximum number of iterations is reached and then stopped. The path with the highest pheromone concentration is selected as the optimal access path, and the finally generated optimized path is passed to the permission application module for dynamic management and implementation of user access requests.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1. The present invention takes the current system state as the root node, and takes the new state generated after taking different actions in the current state as the child node to construct the corresponding exploration tree, calculate the confidence bound value of each node in the exploration tree, and then based on the UCB selection strategy, gradually select the node with the largest confidence bound value until the selected child node is a leaf node that has not been fully visited. According to each group of actions that can be taken under the system state of the current leaf node, a corresponding new state is generated, and the generated new state is added to the exploration tree as a new child node. The benefits of each group of expanded child node states are evaluated through random simulation until the termination state or the preset number of steps is reached, and the simulation is stopped. After the simulation is completed, the simulation results are traced back to all the nodes passed on the path, and the number of visits and cumulative benefits of each node are updated. Selection, expansion, simulation and backtracking update are repeated until the preset iteration time is reached and stopped, all node information in the exploration tree is collected, and the action setting corresponding to the node path with the most visits is selected as the preliminary permission strategy, which can quickly generate candidate permission strategies to meet real-time decision-making needs, and at the same time provide more selection space for subsequent selections, help users more intuitively understand the basis for strategy generation, and help optimize and adjust permission management strategies.

[0042] 2. The present invention constructs all resources and access rights in the system into a graph, wherein the nodes in the graph represent resources, and the edges represent feasible access paths. The cost of the access path is used as the weight of each edge, and the pheromone concentration is initialized on each edge. At the same time, the heuristic value of each group of access paths is calculated according to security and response time, multiple groups of adjustment individuals are initialized, and the positions of the initial groups of adjustment individuals in the graph are randomly distributed. The selection probability of each group of paths in the graph is calculated according to the pheromone concentration and the heuristic value, and the next access path is randomly selected from the starting node according to the selection probability of each path to access the target node in turn until the end point is reached. After the path traversal is completed, the weight pheromone of the edge passed is updated, and the path selection, path construction and pheromone update are repeated until the maximum number of iterations is reached and stopped, and the path with the highest pheromone concentration is selected as the optimal access path, and the finally generated optimized path is passed to the permission application module for dynamic management and implementation of user access requests, which can adjust the access path in real time, improve the flexibility of the system, and perform global search in a complex permission space to avoid local optimality, thereby ensuring the continuity and stability of access. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0044] Figure 1 This is a system block diagram of a data access rights dynamic adjustment system proposed by the present invention. DETAILED DESCRIPTION

[0045] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0046] Example 1

[0047] Reference Figure 1 ,A data access permission dynamic adjustment system includes a data collection module, a feature extraction module, a risk assessment module, a policy generation module, a decision optimization module, a verification and adjustment module, a permission application module, a user feedback module, an omaly detection module, and a monitoring and recording module.

[0048] The data collection module is used to collect user behavior data, access requests and various network environment information in real time; the feature extraction module is used to pre-process the original data collected by the data collection module and perform feature extraction and dimensionality reduction processing.

[0049] Specifically, the missing values, duplicate values ​​and abnormal data in each group of original data are removed, and the data of different dimensions are standardized to the same range through the standard score normalization method. The processed groups of data are input into the pre-trained CNN network. After receiving each group of input data, the CNN network uses groups of convolution kernels of different sizes in the first convolution layer to perform convolution operations on the input data to extract local features, and capture different aspects of the data based on multiple convolution kernels to generate multiple feature maps. The feature maps output by the convolution layer are nonlinearly activated through the ReLU activation function, and each feature map is element-by-element operated to generate a nonlinear feature map. The generated nonlinear feature maps are The image is subjected to the maximum pooling operation through the pooling layer, and the maximum value of each area in the feature map is retained to obtain the downsampled feature map. The output of the pooling layer is then input into the new convolution layer for convolution operation again. The convolution, activation and pooling processes are repeated until the preset processing cycle is reached, and the final feature map is input. After that, the CNN network fully connected layer flattens the multi-dimensional feature map after convolution and pooling into a one-dimensional vector. The output layer receives the processed one-dimensional vector and post-processes it to output the corresponding groups of feature data. After that, the extracted groups of data are projected into a subspace of lower dimensionality through linear transformation, and each group of feature data is normalized through the L2 regularization method.

[0050] The risk assessment module is used to perform risk assessment on the user's current access request.

[0051] Specifically, the user's current access request features are extracted from each group of feature data after dimensionality reduction, and the extracted feature data are integrated into a request feature set. Different risk categories are collected, and based on the number of requests belonging to the risk level in historical data, the initial estimate of each group of risk categories when the request features are not received is calculated. The probability distribution of each feature data in the request feature set under different risk categories is modeled through a multinomial distribution model, and the probability of observing the current feature under each risk category is obtained based on the modeling result. Then, the posterior probability of each risk category being established after observing any feature is calculated according to the Bayesian theorem, the calculated posterior probabilities of each group are collected, and the category with the highest risk is selected as the current risk assessment result. The thresholds of each risk level are dynamically set according to the actual situation and historical data, and then the corresponding strategy is adopted according to the risk level corresponding to the risk category.

[0052] It should be further explained that the risk levels include low risk, medium risk and high risk; the strategies include: if it is low risk, access is allowed; if it is medium risk, a warning is issued or a secondary authentication is required; if it is high risk, access is denied or a security response is triggered immediately.

[0053] The policy generation module is used to simulate and evaluate different policies based on the risk assessment results to generate preliminary permission policies;

[0054] Specifically, according to the extracted groups of feature data, different system state information is obtained. According to the risk assessment results, the system can take various groups of actions in each state, where each action corresponds to an adjustment of the permission policy. Then, the current system state is taken as the root node, and the new state generated after taking different actions in the current state is taken as the child node to construct the corresponding exploration tree, calculate the confidence bound value of each node in the exploration tree, and then based on the UCB selection strategy, gradually select the node with the largest upper confidence bound value until the selected child node is a leaf node that has not been fully visited. According to the various groups of actions that can be taken under the system state of the current leaf node, the corresponding new state is generated, and the generated new state is added to the exploration tree as a new child node. The benefits of each group of expanded child node states are evaluated through random simulation until the termination state or the preset number of steps is reached, and the simulation is stopped. After the simulation is completed, the simulation results are traced back to all the nodes passed on the path, and the number of visits and cumulative benefits of each node are updated. Selection, expansion, simulation and backtracking update are repeated until the preset iteration time is reached and stopped. All node information in the exploration tree is collected, and the action setting corresponding to the node path with the most visits is selected as the preliminary permission policy.

[0055] In this embodiment, the system may take various groups of actions including allowing access, partial access, denying access, and multi-factor application.

[0056] Example 2

[0057] Reference Figure 1 ,A data access permission dynamic adjustment system includes a data collection module, a feature extraction module, a risk assessment module, a policy generation module, a decision optimization module, a verification and adjustment module, a permission application module, a user feedback module, an omaly detection module, and a monitoring and recording module.

[0058] The decision optimization module is used to optimize the preliminary permission strategy and dynamically optimize the access path.

[0059] Specifically, multiple groups of permission policy information are extracted from historical data to construct a corresponding policy space, and the generated preliminary permission policy is added to the policy space. Then, a group of populations are initialized in the policy space by random generation, and each individual in the population represents a group of potential permission policies. The fitness value of each individual is calculated according to the policy security, user experience and system performance indicators, and three individuals with the highest fitness values ​​are selected according to the fitness values. Then, the positions of the remaining individuals in the policy space are updated by simulating hunting behavior. After updating the individual position, check whether the individual position is out of bounds. If the new position of the individual exceeds the set permission policy range, the individual is corrected, and the fitness value selection, position update and correction are repeated until the fitness converges to the preset range and the iteration is stopped. After the iteration stops, the individual with the highest fitness is used as the optimized permission policy, and the finally generated permission policy is passed to the permission application module, and the user access request is managed according to the final permission policy.

[0060] Specifically, all resources and access rights in the system are constructed into a graph, in which nodes in the graph represent resources and edges represent feasible access paths. The cost of the access path is used as the weight of each edge, and the pheromone concentration is initialized on each edge. At the same time, the heuristic value of each group of access paths is calculated based on security and response time, multiple groups of adjustment individuals are initialized, and the positions of the initial groups of adjustment individuals at the starting nodes in the graph are randomly distributed. The selection probability of each group of paths in the graph is calculated according to the pheromone concentration and the heuristic value, and the next access path is randomly selected from the starting node according to the selection probability of each path to visit the target node in turn until the end point is reached. After the path traversal is completed, the weight pheromone of the passed edge is updated, and the path selection, path construction and pheromone update are repeated until the maximum number of iterations is reached and the path with the highest pheromone concentration is selected as the optimal access path. The finally generated optimized path is passed to the permission application module for dynamic management and implementation of user access requests.

[0061] The verification and adjustment module is used to simulate real access requests, verify and adjust policies; the permission application module is used to deploy verified permission policies to the actual data access control system; the user feedback module is used to collect users' opinions and suggestions on the current permission policies in real time, and optimize the permission adjustment policies; the anomaly detection module is used to compare users' historical behaviors with current behavior patterns, and detect possible abnormal access behaviors in real time; the monitoring and recording module is used to record permission adjustment information and access behaviors.

Claims

1. A data access rights dynamic adjustment system, characterized in that: It includes data collection module, feature extraction module, risk assessment module, strategy generation module, decision optimization module, verification and adjustment module, permission application module, user feedback module, anomaly detection module and monitoring and recording module; The data collection module is used to collect user behavior data, access requests and network environment information in real time; The feature extraction module is used to pre-process the raw data collected by the data collection module and perform feature extraction and dimensionality reduction processing; The risk assessment module is used to perform risk assessment on the user's current access request; The policy generation module is used to simulate and evaluate different policies based on the risk assessment results to generate a preliminary authority policy; The decision optimization module is used to optimize the preliminary permission strategy and dynamically optimize the access path; The verification and adjustment module is used to simulate real access requests, verify and adjust policies; The permission application module is used to deploy the verified permission policy into the actual data access control system; The user feedback module is used to collect users' opinions and suggestions on the current permission policy in real time and optimize the permission adjustment policy; The anomaly detection module is used to compare the user's historical behavior and current behavior pattern, and detect possible abnormal access behavior in real time; The monitoring and recording module is used to record permission adjustment information and access behavior.

2. A data access rights dynamic adjustment system according to claim 1, characterized in that: The specific steps of extracting dimension reduction by the feature extraction module are as follows: S1.1: Remove missing values, duplicate values, and abnormal data from each group of raw data, normalize data of different dimensions to the same range through the standard score normalization method, and input each group of processed data into the pre-trained CNN network; S1.2: After receiving each set of input data, the CNN network uses different sizes of convolution kernels in the first convolution layer to perform convolution operations on the input data to extract local features, and capture different aspects of the data based on multiple convolution kernels to generate multiple feature maps. The feature maps output by the convolution layer are nonlinearly activated through the ReLU activation function, and each feature map is element-by-element operated to generate a nonlinear feature map; S1.3: The generated nonlinear feature map is passed through the pooling layer for maximum pooling operation, and the maximum value of each area in the feature map is retained to obtain the downsampled feature map. The output of the pooling layer is then input into the new convolution layer for convolution operation again. The convolution, activation and pooling processes are repeated until the preset processing cycle is reached. The final feature map is input, and then the CNN network fully connected layer flattens the multi-dimensional feature map after convolution and pooling into a one-dimensional vector. S1.4: The output layer receives the processed one-dimensional vector and post-processes it to output the corresponding groups of feature data. Then, each group of extracted data is projected into a subspace with lower dimensionality through linear transformation, and each group of feature data is normalized through L2 regularization method.

3. A data access rights dynamic adjustment system according to claim 2, characterized in that: The specific steps of the risk assessment module performing risk assessment on the user's current access request are as follows: S2.1: Extract the user's current access request features from each group of feature data after dimensionality reduction, integrate the extracted feature data into a request feature set, collect different risk categories, and calculate the initial estimate of each group of risk categories when no request features are received based on the number of requests belonging to the risk level in the historical data; S2.2: The probability distribution of each feature data in the requested feature set under different risk categories is modeled through a multinomial distribution model, and the probability of observing the current feature under each risk category is obtained based on the modeling results. Then, the posterior probability of each risk category being established after observing any feature is calculated based on the Bayesian theorem; S2.3: Collect the calculated posterior probabilities of each group and select the category with the highest risk as the current risk assessment result. Dynamically set the thresholds of each risk level based on the actual situation and historical data. Then, adopt the corresponding strategy based on the risk level corresponding to the risk category.

4. A data access rights dynamic adjustment system according to claim 3, characterized in that: The specific steps of the policy generation module generating a preliminary authority policy are as follows: S3.1: Based on the extracted feature data, different system status information is obtained. Based on the risk assessment results, the system can take different sets of actions in each state, where each action corresponds to an authority policy adjustment. The current system state is then used as the root node, and the new states generated after taking different actions in the current state are used as child nodes to construct the corresponding exploration tree. S3.2: Calculate the confidence bound value of each node in the exploration tree, and then based on the UCB selection strategy, gradually select the node with the largest confidence bound value until the selected child node is a leaf node that has not been fully visited. Generate the corresponding new state according to the various groups of actions that can be taken under the system state of the current leaf node, and add the generated new state as a new child node to the exploration tree; S3.3: Evaluate the benefits of each group of expanded child nodes through random simulation until the terminal state or the preset number of steps is reached, then stop the simulation. After the simulation is completed, trace the simulation results back to all the nodes passed on the path, and update the number of visits and cumulative benefits of each node; S3.4: Repeat selection, expansion, simulation, and backtracking until the preset iteration time is reached, collect all node information in the exploration tree, and select the action setting corresponding to the node path with the most visits as the preliminary permission strategy.

5. A data access rights dynamic adjustment system according to claim 4, characterized in that: The specific steps of optimizing the preliminary authority strategy by the decision optimization module are as follows: S4.1: Extract multiple groups of permission policy information from historical data to construct a corresponding policy space, and add the generated preliminary permission policies to the policy space. Then, initialize a population in the policy space by random generation, and each individual in the population represents a group of potential permission policies. S4.2: Calculate the fitness value of each individual based on the security, user experience and system performance indicators of the strategy, select the three individuals with the highest fitness values ​​based on the fitness values, and then update the positions of the remaining individuals in the strategy space by simulating hunting behavior; S4.3: After updating the individual position, check whether the individual position is out of bounds. If the new position of the individual exceeds the set authority policy range, the individual is corrected, and the fitness value selection, position update and correction are repeated until the fitness converges to the preset range and the iteration is stopped; S4.4: After the iteration stops, the individual with the highest fitness is used as the optimized permission strategy, and the final permission strategy is passed to the permission application module, and the user access request is managed according to the final permission strategy.

6. A data access rights dynamic adjustment system according to claim 1, characterized in that: The specific steps of dynamically optimizing the access path by the decision optimization module are as follows: S5.1: Construct all resources and access rights in the system into a graph, where nodes represent resources and edges represent feasible access paths. The cost of the access path is used as the weight of each edge, and the pheromone concentration is initialized on each edge. At the same time, the heuristic value of each group of access paths is calculated based on security and response time. S5.2: Initialize multiple groups of adjustment individuals, and randomly distribute the positions of the initial groups of adjustment individuals at the starting nodes in the graph, calculate the selection probability of each group of paths in the graph according to the pheromone concentration and the heuristic value, and randomly select the next access path from the starting node according to the selection probability of each path to visit the target node in sequence until the end point is reached; S5.3: After the path traversal is completed, the weight pheromones of the edges passed are updated, and the path selection, path construction and pheromone update are repeated until the maximum number of iterations is reached and then stopped. The path with the highest pheromone concentration is selected as the optimal access path, and the finally generated optimized path is passed to the permission application module for dynamic management and implementation of user access requests.

Images