Scenario-adaptive permission dynamic adjustment method and device based on trust evaluation

Abstract

The present invention discloses a method and device for dynamically adjusting scene-adaptive permissions based on trust evaluation, the method comprising: obtaining an access request initiated by a user, obtaining the role permissions corresponding to the user based on the access request; obtaining the context conditions corresponding to the access request, determining the access control policy based on the context conditions, and adjusting the role permissions corresponding to the user based on the access control policy; performing trust evaluation on the user based on a pre-established trust evaluation model and a trust threshold to obtain a trust evaluation result, wherein the trust evaluation model is established based on the user's historical behavior through a fuzzy Q learning method; and performing permission management on the access request based on the adjusted role permissions corresponding to the user and the trust evaluation result. The present invention solves the technical problem in the prior art that it is impossible to accurately evaluate in advance whether the user's behavior is legal.

Description

Technical Field

[0001] The present invention relates to the field of data security technology, and in particular to a method, device, electronic device and storage medium for dynamically adjusting scene-adaptive permissions based on trust evaluation. Background Art

[0002] As digital transformation continues to accelerate, emerging technologies and innovative businesses continue to break the original security boundaries, and information security faces unprecedented challenges. In the static traditional environment, the access control of the system is mostly based on the subject-object, passive security model. The static nature of their authorization policies determines that from the beginning of the user's authorization to the end of the access operation, unless the administrator actively revokes the user's permissions, the user's granted permissions can continue to have them. In addition, the authorization context is not taken into account, and the authorization system has certain hidden dangers. However, in the current complex network environment, due to the mobility of devices, traditional access control strategies are no longer suitable for providing complete security for network resources. The dynamic environment in which the entity is located is the most important factor in the adaptive access control strategy. In order to ensure the security of devices in a dynamic environment, it is necessary to study efficient, flexible, and scalable scenario-adaptive access control.

[0003] Most traditional access control models are based on static access control policies, which cannot meet the dynamic and fine-grained access control requirements in open network environments. To address this problem, many researchers have proposed access control based on zero-trust architecture. Zero trust is a security model whose core principle is "continuous verification, never trust". All users are continuously verified and dynamically authorized based on as many trust factors as possible, such as the identity of the access subject, the network environment, and the terminal status. Some researchers have proposed a dynamic access control and authorization system based on a zero-trust security architecture, which generates user profiles and user trust through user behavior, and uses real-time hierarchical control in the system to achieve dynamic and fine-grained access control and authorization. At the same time, some researchers have proposed two methods to improve trust-based access control, one is attribute weight evaluation based on information entropy, and the other is recommendation trust evaluation based on grey correlation analysis.

[0004] The limitations of existing solutions are mainly reflected in:

[0005] (1) Existing work often lacks comprehensive context awareness when evaluating user behavior. For example, it may only consider the user's identity and permissions, while ignoring other contextual information related to user behavior, such as device, location, time, etc. This may lead to an inability to accurately assess whether the user's behavior is legal.

[0006] (2) Most existing solutions rely on post-detection to detect illegal user behavior. This means that measures can only be taken to respond after the user behavior occurs. Since illegal behavior cannot be predicted in advance, once illegal behavior occurs, it may have already caused damage to the system. Summary of the invention

[0007] The purpose of the present invention is to overcome the above-mentioned technical deficiencies and provide a scenario-adaptive permission dynamic adjustment method, device, electronic device and storage medium based on trust evaluation to solve the technical problem in the prior art that it is impossible to accurately evaluate in advance whether the user's behavior is legal.

[0008] In order to achieve the above technical objectives, the present invention adopts the following technical solutions:

[0009] In a first aspect, the present invention provides a scenario-adaptive permission dynamic adjustment method based on trust evaluation, comprising the following steps:

[0010] Obtaining an access request initiated by a user, and obtaining the role authority corresponding to the user based on the access request;

[0011] Obtaining a context condition corresponding to the access request, determining an access control policy based on the context condition, and adjusting the role authority corresponding to the user based on the access control policy;

[0012] Based on a pre-established trust evaluation model and a trust threshold, a trust evaluation is performed on the user to obtain a trust evaluation result, wherein the trust evaluation model is established based on the user's historical behavior through a fuzzy Q learning method;

[0013] Based on the adjusted role permissions corresponding to the user and the trust evaluation result, permission management is performed on the access request.

[0014] In some embodiments, obtaining an access request initiated by a user and obtaining role permissions corresponding to the user based on the access request includes:

[0015] Obtaining an access request initiated by a user, and parsing the access request to obtain user information contained in the access request;

[0016] Based on the user information, the role authority corresponding to the user is obtained.

[0017] In some embodiments, the contextual conditions include at least a time context, a spatial context, a user-declared context, and a prerequisite context. The time context is determined based on an access time of the access request, the spatial context is determined based on a location of the access request, the user-declared context is determined based on an access purpose of the access request, and the prerequisite context is determined based on the access request and a server database.

[0018] In some embodiments, obtaining the context condition corresponding to the access request, determining the access control policy based on the context condition, and adjusting the role authority corresponding to the user based on the access control policy includes:

[0019] Based on the access request, obtaining a context condition corresponding to the access request;

[0020] Based on the context condition, obtaining an access control policy corresponding to the context condition, wherein the access policy is one or more of a rule, a specification or a logical expression;

[0021] Based on the access control policy, the role permissions corresponding to the user are adjusted.

[0022] In some embodiments, the trust evaluation of the user based on the pre-established trust evaluation model and trust threshold to obtain a trust evaluation result includes:

[0023] Based on the access request, obtaining a trust threshold corresponding to the access request;

[0024] Calculate the trust evaluation score corresponding to the access request based on a pre-established trust evaluation model;

[0025] A trust evaluation result is determined based on the trust evaluation score and the trust threshold.

[0026] In some embodiments, the method for establishing the trust assessment model is:

[0027] Step 1: Establish a fuzzy inference system rule base, wherein the fuzzy inference system rule base consists of N rules;

[0028] Step 2: Obtain the user's historical access records to obtain the user's current status, and perform fuzzy processing on the user's current status, where the status is a five-tuple consisting of user ID, role, requested resource, permission granted, and context conditions;

[0029] Step 3: Based on the current state of the user after the fuzzification process, for each triggered rule corresponding to the access request, randomly select the action of the triggered rule according to the ε-greedy strategy to calculate the action of the triggered rule and its corresponding Q value;

[0030] Step 4: Based on the actions of the triggered rules and their corresponding Q values, calculate the global actions and their corresponding Q values;

[0031] Step 5: executing the calculated global action and updating the current state, and after calculating the reward value, updating the result vector weight of the fuzzy inference system rule base according to the reward value obtained by the global action;

[0032] Step 6: Repeat steps 2 to 5 until the result vector weights converge, so as to generate a trust assessment model according to the fuzzy inference system rule base.

[0033] In some embodiments, the calculation formula of the global action and its corresponding Q value is:

[0034]

[0035]

[0036] Among them, a(s) represents the global action, Q(s,a) represents the Q value corresponding to the global action, and w i (x) represents the trigger strength of rule i, q[i,i ′ ] represents the i-th ′ The Q value corresponding to the possible action is a i represents the action of rule i, and N represents the number of rules.

[0037] In a second aspect, the present invention further provides a scenario-adaptive authority dynamic adjustment device based on trust evaluation, comprising:

[0038] The permission acquisition module is used to obtain the access request initiated by the user and obtain the role permission corresponding to the user based on the access request;

[0039] A permission adjustment module, used to obtain the context conditions corresponding to the access request, determine the access control policy based on the context conditions, and adjust the role permissions corresponding to the user based on the access control policy;

[0040] A trust evaluation module, used to perform trust evaluation on the user based on a pre-established trust evaluation model and a trust threshold to obtain a trust evaluation result, wherein the trust evaluation model is established based on the user's historical behavior through a fuzzy Q learning method;

[0041] The permission granting module is used to perform permission management on the access request based on the adjusted role permission corresponding to the user and the trust evaluation result.

[0042] In a third aspect, the present invention further provides an electronic device, comprising: a processor and a memory;

[0043] The memory stores a computer program executable by the processor;

[0044] When the processor executes the computer program, the steps in the scenario-adaptive authority dynamic adjustment method based on trust evaluation as described above are implemented.

[0045] In a fourth aspect, the present invention further provides a computer-readable storage medium, comprising: a processor and a memory;

[0046] The memory stores a computer program executable by the processor;

[0047] When the processor executes the computer program, the steps in the scenario-adaptive authority dynamic adjustment method based on trust evaluation as described above are implemented.

[0048] Compared with the prior art, the scenario-adaptive permission dynamic adjustment method, device, electronic device and storage medium provided by the present invention based on trust evaluation have the following beneficial effects:

[0049] (1) Strong adaptability to scenarios: The present invention fully considers the contextual conditions of access scenarios and can adjust permissions in a timely manner according to environmental changes to achieve more fine-grained authorization and help the system better protect sensitive information and resources. For example, in scenarios such as medical information systems or logistics management systems where patient flow and logistics flow are changeable, the present invention can fully consider contextual conditions such as patient department flow or logistics node changes to make corresponding permission adjustments;

[0050] (2) Flexible and adjustable: The present invention uses a trust assessment strategy based on fuzzy Q learning, collects real-time data for learning, and uses fuzzy Q learning methods to analyze and process data. It can select the optimal action based on the current state of the environment and the behavior of the subject. By continuously learning and updating the trust model, the strategy can adapt to different contexts and risk situations. In different scenarios, flexible trust assessments based on real-time data and environmental conditions support permission adjustments. Especially in scenarios with high data security requirements such as medical information systems or logistics management systems, the strategy can evaluate the user's trust level based on their behavior and historical data, and restrict or open their access rights accordingly;

[0051] (3) Safe and controllable: While providing dynamic permission adjustment, the present invention establishes a trust evaluation mechanism based on user historical access records. By analyzing user historical access records and behavior patterns, it can actively block illegal requests from seemingly legitimate users. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is a flow chart of a scenario-adaptive authority dynamic adjustment method based on trust evaluation provided by an embodiment of the present invention;

[0053] Figure 2 It is a flowchart of step S100 in the scenario-adaptive authority dynamic adjustment method based on trust evaluation provided in an embodiment of the present invention;

[0054] Figure 3 It is a flowchart of step S200 in the scenario-adaptive authority dynamic adjustment method based on trust evaluation provided in an embodiment of the present invention;

[0055] Figure 4 A schematic diagram of a trust evaluation model in a scenario-adaptive permission dynamic adjustment method based on trust evaluation provided in an embodiment of the present invention;

[0056] Figure 5 It is a flowchart of step S300 in the scenario-adaptive authority dynamic adjustment method based on trust evaluation provided in an embodiment of the present invention;

[0057] Figure 6 It is a functional module diagram of a scene-adaptive authority dynamic adjustment device based on trust evaluation provided by an embodiment of the present invention;

[0058] Figure 7 It is a schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0060] The following first explains and illustrates the technical terms of the present invention:

[0061] Contextual conditions: Any information about the environment itself and the entities in the environment that can be used to describe their state (including historical state), either explicitly or implicitly. Entities can be physical entities such as people and places, or virtual entities such as software, programs, and network connections.

[0062] Subject: Mainly refers to the user who initiates the request.

[0063] Objects: include data, documents, and all data sources and computing services considered as resources in the system.

[0064] Trust value: quantifies the degree of trust that the system has in the subject.

[0065] Trust threshold: Indicates the minimum trust value required to operate a resource.

[0066] See also Figure 1 The method for dynamically adjusting scene-adaptive permissions based on trust evaluation provided by the present invention comprises the following steps:

[0067] S100, obtaining an access request initiated by a user, and obtaining the role authority corresponding to the user based on the access request;

[0068] S200, obtaining a context condition corresponding to the access request, determining an access control policy based on the context condition, and adjusting the role authority corresponding to the user based on the access control policy;

[0069] S300, based on a pre-established trust evaluation model and a trust threshold, performing a trust evaluation on the user to obtain a trust evaluation result, wherein the trust evaluation model is established based on the user's historical behavior through a fuzzy Q learning method;

[0070] S400: Performing authority management on the access request based on the adjusted role authority corresponding to the user and the trust evaluation result.

[0071] In this embodiment, an initial role is first assigned to each user through role allocation, and then the dynamic environment in which the user is located is abstractly modeled, and its contextual condition information is monitored, and its permissions are dynamically adjusted on a role-based basis. Then, a trust assessment model is built according to the user's historical access records and current access status, and the user's trust is evaluated. Finally, the user's permissions are managed according to the user's current access behavior status, the trust assessment results, and the adjusted role permissions.

[0072] On the one hand, the embodiment of the present invention continuously evaluates the user behavior status by means of fuzzy reinforcement learning, thus avoiding the threat of a legitimate user suddenly engaging in illegal behavior after a single authentication. On the other hand, the context information of the complex dynamic network environment is fully considered, and the user access environment is included in the decision-making conditions for granting permissions, thereby preventing access to sensitive information in an unsafe environment. In addition, the context condition constraints and trust evaluation are fully combined to monitor the entire process of user access behavior and dynamically adjust the permissions applicable to the scenario in a timely manner based on the relevant status information.

[0073] In addition, the embodiment of the present invention has taken into account the dynamic adjustment strategy of data permissions in a fine-grained manner, including the contextual conditions of the current access environment and the current access status and historical access status, and has monitored the access behavior throughout the entire process to ensure the scenario-adaptive adjustment of permissions. The trust assessment model in this method regards all access users as untrusted users, and assigns access levels based on their historical access records and current access status, thereby limiting the upper limit of their access rights, avoiding legal users from suddenly initiating illegal access behaviors, and realizing dynamic adjustment of permissions based on trust assessment.

[0074] It is understandable that the solution provided by the embodiment of the present invention is applicable to the information systems of modern enterprises and organizations, which contain many users and also have remote access by users. The solution provided by the embodiment of the present invention can grant permissions in a timely manner based on the different roles of users, access scenarios and previous access records.

[0075] In some embodiments, in step S100, except for the system administrator, all other user roles are assigned by the system administrator. The role enables the system administrator to bundle a set of permissions to one role. The system administrator can also create custom roles as needed. The role limits the highest permission possessed by the subject and also determines, to a certain extent, the worst case scenario in which the subject can violate the regulations.

[0076] Optional, see Figure 2 , the step S100 specifically includes:

[0077] S110, obtaining an access request initiated by a user, and parsing the access request to obtain user information included in the access request;

[0078] S120: Based on the user information, obtain the role authority corresponding to the user.

[0079] In this embodiment, when a user initiates an access request, the user information corresponding to the user is sent to the server at the same time. The server parses the access request, obtains the user information contained in the access request, and then finds the role corresponding to the user based on the user information, and then obtains the role permissions corresponding to the user.

[0080] In some embodiments, step S200 is performed by evaluating the access environment attributes of the subject O and representing them through a set of context condition parameters CC. When the subject requests the operation permission for the object S, the permission is constrained based on the context information of the entity based on the permission of the role possessed by the user, so as to realize dynamic adjustment of the permission.

[0081] Optionally, in step S200, the context conditions include at least time context, spatial context, user-declared context and prerequisite context, the time context is determined based on the access time of the access request, the spatial context is determined based on the location of the access request, the user-declared context is determined based on the access purpose of the access request, and the prerequisite context is determined based on the access request and a server database.

[0082] In this embodiment, the context condition in the information system is defined as CC = {CCt, CCs, CCg, CCp}, which respectively represent the time context, space context, user-declared context, and prerequisite context. In order to control the context activation, each system provides the information needed to verify whether it is consistent with or different from the context definition. The context condition refers to the environment itself and any information explicitly or implicitly used to describe its state (including historical state) by each entity in the environment. Among them, the time context depends on the access time of the subject requesting the system, that is, the global clock is used to verify the time context; the space context depends on the location of the subject, that is, the environment of the subject and the architecture of software and hardware are used to verify the space context; the user-declared context depends on the purpose of the subject; the prerequisite context depends on the characteristics of the connection subject, action and object, that is, the system database is used to verify the prerequisite context.

[0083] Optional, see Figure 3 , the step S200 specifically includes:

[0084] S210, based on the access request, obtaining a context condition corresponding to the access request;

[0085] S220, based on the context condition, obtaining an access control policy corresponding to the context condition, where the access policy is one or more of a rule, a specification, or a logical expression;

[0086] S230: Based on the access control policy, adjust the role authority corresponding to the user.

[0087] In this embodiment, first, the context condition is identified according to the access request, and the system defines the corresponding permission policy for each context condition type. These policies may include rules, specifications, or logical expressions for adjusting permissions according to the context conditions. For example, for the time context condition, specific permissions within a certain time period may be defined, and the corresponding permission adjustment mechanism may be implemented according to the defined permission policy.

[0088] It is understandable that during the user's access process, the corresponding context conditions are always monitored. When changes occur, the permissions are dynamically adjusted according to the corresponding access control policy to ensure the user's access security throughout the process.

[0089] In some embodiments, step S300 is to abstract the trust assessment engine into an intelligent agent, and abstract the authority decision process of the subject into a Markov decision process. The system uses fuzzy Q learning to establish a trust assessment model based on the user's historical behavior, such as Figure 4 As shown in the figure, the best permission adjustment action can be made according to the current state. Specifically, it includes obtaining the current subject state, selecting a permission granting action to execute through the fuzzy reasoning system, and using the Q algorithm to continuously update the result vector weight in the rule base until it converges. Finally, a complete rule base is obtained to provide prior knowledge for the behavior selection of the trust assessment engine.

[0090] Optional, see Figure 5 , the step S300 specifically includes:

[0091] S310: Based on the access request, obtain a trust threshold corresponding to the access request;

[0092] S320: Calculate a trust evaluation score corresponding to the access request based on a pre-established trust evaluation model;

[0093] S330: Determine a trust evaluation result based on the trust evaluation score and the trust threshold.

[0094] In this embodiment, the trust value and trust threshold are initialized first. Specifically, all subjects with the same role have the same trust value when they access for the first time. The trust threshold of the object is initialized according to the importance and security level of the resource, and it is also dynamically adjusted according to changes in the environment where the resource is located. For example, the importance and security level of the resource may decrease over time. This lowers the trust threshold of the resource, allowing users to operate the resource with a lower behavioral trust. Afterwards, the trust assessment score corresponding to the access request is calculated through the trust assessment model, and the trust assessment score is compared with the trust threshold. If the trust assessment score is greater than the trust threshold, it means that authorization can be obtained, otherwise it means that further identity confirmation is required.

[0095] Optionally, the trust evaluation model is established by:

[0096] Step 1: Establish a fuzzy inference system rule base, wherein the fuzzy inference system rule base consists of N rules;

[0097] Step 2: Obtain the user's historical access records to obtain the user's current status, and perform fuzzy processing on the user's current status, where the status is a five-tuple consisting of user ID, role, requested resource, permission granted, and context conditions;

[0098] Step 3: Based on the current state of the user after the fuzzification process, for each triggered rule corresponding to the access request, randomly select the action of the triggered rule according to the ε-greedy strategy to calculate the action of the triggered rule and its corresponding Q value;

[0099] Step 4: Based on the actions of the triggered rules and their corresponding Q values, calculate the global actions and their corresponding Q values;

[0100] Step 5: executing the calculated global action and updating the current state, and after calculating the reward value, updating the result vector weight of the fuzzy inference system rule base according to the reward value obtained by the global action;

[0101] Step 6: Repeat steps 2 to 5 until the result vector weights converge, so as to generate a trust assessment model according to the fuzzy inference system rule base.

[0102] In this embodiment, first, the fuzzy inference system rules are defined. The fuzzy inference system rule base consists of N rules.

[0103] if s is S i then a[i,1]with q[i,1]

[0104] or a[i,j]with q[i,j] ...

[0105] or a[i,k]with q[i,k]

[0106] Where a[i,k] is the kth possible action in rule i, and q[i,k] is the corresponding Q value. i is defined as (x1is S i,1 and x2is S i,2 …and x n is S i,n ), where S i,j (i=1,…,nj=1,…,n) is a fuzzy set.

[0107] Then, the current state s is obtained. Specifically, the embodiment of the present invention defines the historical access record of the subject as state s = {id, role, source, grant, CC}, which respectively represent the user ID, role, requested resource, permission granted, and the aforementioned context conditions. The permission granting action a = {Y, D, N}, which respectively represent normal authorization permission, reduced permission, and denied authorization. Reward function grant=1 indicates normal granting of permissions, and grant=-1 indicates abnormal granting of permissions.

[0108] Then, an action is selected. For each triggered rule, an action is randomly selected according to the ε-greedy strategy, and then the global action a(s) and the corresponding Q(s, a) are calculated. Specifically, the fuzzy inference system is defuzzified by the weighted average method.

[0109] Optionally, the calculation formula of the global action and its corresponding Q value is:

[0110]

[0111]

[0112] Among them, a(s) represents the global action, Q(s,a) represents the Q value corresponding to the global action, and w i (x) represents the trigger strength of rule i, q[i,i ′ ] represents the i-th ′ The Q value corresponding to the possible action is a i represents the action of rule i, and N represents the number of rules.

[0113] Finally, execute action a(s) and obtain the new state s ′ , and calculate the reward value, and update the result vector weight of the rule base according to the reward value obtained by the action.

[0114] It is understandable that in order to enhance the security performance of the system, the trust assessment model needs to repeatedly execute steps 2 to 5 until the vector weights converge. And the trust assessment model can be optimized regularly in an incremental learning manner based on new historical access records.

[0115] In some embodiments, step S400 is the final permission management step. Specifically, permission is a binary group consisting of an operation and an object.<Operate,Subject> . Operation: The operation set includes read, write, execute, store, publish, forward (the behavior of a common subject forwarding all or part of the published data to another subject), and select (the operation used by the subject to select the data range to be operated). These 7 operations are atomic, and most operations can be obtained by combining these atomic operations. The embodiment of the present invention constrains the subject's permissions according to contextual conditions, and decides whether to authorize or require further identity authentication according to the trust evaluation results.

[0116] In order to better understand the present invention, a specific embodiment is given below to describe the technical solution of the present invention in detail:

[0117] In this embodiment, the scenario of the use of the scenario-adaptive permission dynamic adjustment method based on trust assessment is an electronic medical system. The access control under this system has higher requirements for the security of access control and data confidentiality than other fields, and the permission control requirements in the medical system are very complex, involving multiple roles, multiple departments and different patient statuses. It is necessary to accurately divide permissions based on factors such as the doctor's or nurse's position, qualifications, department affiliation and the patient's privacy needs. At the same time, the patient's status may change at any time, such as transfer, discharge, etc., and it is necessary to monitor in real time and adjust the doctor's or nurse's access rights accordingly to ensure that only the right people can access the corresponding patient information. Using fuzzy reinforcement learning can be dynamically adjusted according to actual conditions and environmental changes. When the patient transfers, is discharged or other status changes, the system can monitor in real time and adjust the doctor's or nurse's access rights accordingly to ensure the confidentiality and security of the data. This adaptability and flexibility enable the system to adapt to complex hospital environments and dynamic changes in patient files. The specific steps are as follows:

[0118] (1) Role assignment: Except for the system administrator, all other user roles are assigned by the administrator. The system administrator can bind a set of permissions to a role. The administrator can also create custom roles as needed. The role limits the highest permissions that the subject has and also determines the worst case scenario in which the subject can violate the rules. In the electronic medical system, there are many departments and medical staff. Medical staff can access the patient information of their own department based on their own identity and cannot access the information of any other department.

[0119] (2) Contextual condition constraints: The access environment attributes of the subject O are evaluated and represented by a set of contextual condition parameters CC. When the subject requests the operation permission for the object S, we constrain the permission based on the context information of the entity based on the permissions of the role owned by the user to achieve dynamic adjustment of the permissions. When a patient needs to be transferred to another department, the system will update the department information of the patient file and adjust the permissions of the doctor or nurse accordingly. For example, if Mr. Zhang needs to be transferred to the cardiology department, the system will update the department information of Mr. Zhang's patient file to the cardiology department and assign it to the doctor or nurse of the cardiology department for management. The doctor or nurse of the original department has no permission to access the patient file information. It includes the following steps:

[0120] (2.1) Determine the context type: We define the context condition in the information system as CC = {CCt, CCs, CCg, CCp}, which represents the time context, space context, user declaration context, and prerequisite context, respectively. The time context refers to the access time, the space context refers to the access IP and device, and the prerequisite context refers to the attribute information of the patient itself, such as admission, transfer to another department, discharge, etc. Different changes in conditions will affect the granting of access rights.

[0121] (2.2) Identify contextual conditions: In order to control context activation, each system provides the information needed to check whether the situation is consistent with or different from the context definition. A global clock is used to check the time context; the subject's environment and the software and hardware architecture are used to check the spatial context; the subject's purpose is used to check the user declaration context; the system's database is used to check the prerequisite context. The identification of contextual conditions needs to be activated at the beginning of each access, and the changes in the upper and lower conditions are continuously monitored to update the contextual conditions in a timely manner.

[0122] (2.3) Develop access control policies: For each context condition type, define the corresponding permission policy. Constrain based on the context conditions based on the role-based permissions. These policies can include rules, specifications, or logical expressions to adjust permissions based on the context conditions. For example, for the time context condition, you can define specific permissions within a certain time period.

[0123] (2.4) Implement the permission adjustment mechanism: Implement the corresponding permission adjustment mechanism according to the defined permission strategy. During the user's access process, the corresponding context conditions are always monitored. When changes occur, the permissions are dynamically adjusted according to the corresponding access control strategy. When a patient needs to be transferred to another department, the system will update the department information of the patient file and adjust the doctor or nurse's permissions accordingly. For example, if Mr. Zhang needs to be transferred to the cardiology department, the system will update the department information of Mr. Zhang's patient file to the cardiology department and assign it to the doctor or nurse of the cardiology department for management. The doctor or nurse of the original department will not have the permission to access the patient file information.

[0124] (3) Trust assessment model establishment: The trust assessment engine is abstracted into an intelligent agent, and the authority decision-making process of doctors or nurses is abstracted into a Markov decision process. The system uses fuzzy Q learning to establish a trust assessment model based on the historical behavior of the subject, so as to make the best authority adjustment action based on the current state. Specifically, it includes obtaining the current subject state, selecting an authority granting action to execute through the fuzzy reasoning system, and using the Q algorithm to continuously update the result vector weights in the rule base until it converges. Finally, a complete rule base is obtained to provide prior knowledge for the behavior selection of the trust assessment engine.

[0125] (3.1) Initialize trust value and trust threshold: Different trust values ​​are assigned to doctors and nurses at the first visit. The trust threshold of the object is initialized according to the importance and security level of the resource, and is also dynamically adjusted according to changes in the environment where the resource is located. For example, patient personal information requires a higher trust threshold, while the trust threshold of patient medical records is lower. At the same time, the importance and security level of resources may decrease over time. This reduces the trust threshold of the resource, allowing users to operate the resource with a lower degree of behavioral trust.

[0126] (3.2) Define the fuzzy inference system rules. The fuzzy inference system rule base consists of N rules.

[0127] if s is S i then a[i,1]with q[i,1]

[0128] or a[i,j]with q[i,j] ...

[0129] or a[i,k]with q[i,k]

[0130] Where a[i,k] is the kth possible action in rule i, such as the trust rating granted, and q[i,k] is the corresponding Q value. i is defined as (x1is S i,1 and x2is Si,2 …and x n is S i,n ), where S i,j (i=1,…,nj=1,…,n) is a fuzzy set.

[0131] (3.3) Get the current state s: We define the subject's historical access record as state s = {id, role, source, grant, CC}, which represents the user ID, role, requested resource, permission granted, and the aforementioned context conditions. Permission grant action a = {Y, D, N}, which represents normal authorization, reduced authorization, and denied authorization. Reward function grant=1 indicates normal granting of permissions, and grant=-1 indicates abnormal granting of permissions.

[0132] (3.4) Action selection: For each triggered rule, an action is randomly selected according to the ε-greedy strategy.

[0133] (3.5) Calculate the global action a(s) and the corresponding Q(s,a): defuzzify it by weighted averaging method according to the fuzzy inference system.

[0134]

[0135]

[0136] where w i (x) represents the trigger strength of rule i. i is the action of rule i, and N is the number of rules.

[0137] (3.6) Execute action a(s) and obtain the new state s ′ , and calculate the reward value. Update the result vector weight of the rule base according to the reward value obtained by the action.

[0138] Steps (3.3)-(3.6) are repeated until the vector weights converge. After that, the permission granting action with the maximum weight can be selected for the doctor or nurse according to the access status.

[0139] (3.7) Trust assessment model update: The trust assessment model is updated regularly through incremental learning based on new historical access records.

[0140] (4) Permission grant: Permission is a pair consisting of an operation and an object.<Operate,Subject> Operation: The operation set includes read, write, execute, store, publish, forward (the behavior of a common subject forwarding all or part of the published data to another subject), and select (the operation used by the subject to select the data range to be operated). These 7 operations are atomic, and most operations can be obtained by combining these atomic operations. Constrain the subject's permissions according to contextual conditions, and decide whether to authorize or require further identity authentication based on the trust evaluation results.

[0141] Another embodiment of the present invention provides a scenario-adaptive permission dynamic adjustment device based on trust evaluation, see Figure 6 The scenario-adaptive permission dynamic adjustment device based on trust evaluation includes a permission acquisition module 11, a permission adjustment module 12, a trust evaluation module 13 and a permission granting module 14.

[0142] The authority acquisition module 11 is used to acquire an access request initiated by a user, and acquire the role authority corresponding to the user based on the access request.

[0143] The permission adjustment module 12 is used to obtain the context conditions corresponding to the access request, determine the access control policy based on the context conditions, and then adjust the role permissions corresponding to the user based on the access control policy.

[0144] The trust evaluation module 13 is used to perform trust evaluation on the user based on a pre-established trust evaluation model and a trust threshold to obtain a trust evaluation result, wherein the trust evaluation model is established based on the user's historical behavior through a fuzzy Q learning method.

[0145] The authority granting module 14 is used to perform authority management on the access request based on the adjusted role authority corresponding to the user and the trust evaluation result.

[0146] It should be noted that the module referred to in the present invention refers to a series of computer program instruction segments that can perform specific functions, which is more suitable for describing the execution process of dynamic adjustment of scenario-adaptive permissions based on trust assessment than a program. For the specific implementation methods of each module, please refer to the corresponding method embodiments above, which will not be repeated here.

[0147] In some embodiments, the permission acquisition module 11 is specifically used to:

[0148] Obtaining an access request initiated by a user, and parsing the access request to obtain user information contained in the access request;

[0149] Based on the user information, the role authority corresponding to the user is obtained.

[0150] In some embodiments, the contextual conditions include at least a time context, a spatial context, a user-declared context, and a prerequisite context. The time context is determined based on an access time of the access request, the spatial context is determined based on a location of the access request, the user-declared context is determined based on an access purpose of the access request, and the prerequisite context is determined based on the access request and a server database.

[0151] In some embodiments, the permission adjustment module 12 is specifically used to:

[0152] Based on the access request, obtaining a context condition corresponding to the access request;

[0153] Based on the context condition, obtaining an access control policy corresponding to the context condition, wherein the access policy is one or more of a rule, a specification or a logical expression;

[0154] Based on the access control policy, the role permissions corresponding to the user are adjusted.

[0155] In some embodiments, the trust assessment module 13 is specifically used to:

[0156] Based on the access request, obtaining a trust threshold corresponding to the access request;

[0157] Calculate the trust evaluation score corresponding to the access request based on a pre-established trust evaluation model;

[0158] A trust evaluation result is determined based on the trust evaluation score and the trust threshold.

[0159] In some embodiments, the method for establishing the trust assessment model is:

[0160] Step 1: Establish a fuzzy inference system rule base, wherein the fuzzy inference system rule base consists of N rules;

[0161] Step 2: Obtain the user's historical access records to obtain the user's current status, and perform fuzzy processing on the user's current status, where the status is a five-tuple consisting of user ID, role, requested resource, permission granted, and context conditions;

[0162] Step 3: Based on the current state of the user after the fuzzification process, for each triggered rule corresponding to the access request, randomly select the action of the triggered rule according to the ε-greedy strategy to calculate the action of the triggered rule and its corresponding Q value;

[0163] Step 4: Based on the actions of the triggered rules and their corresponding Q values, calculate the global actions and their corresponding Q values;

[0164] Step 5: executing the calculated global action and updating the current state, and after calculating the reward value, updating the result vector weight of the fuzzy inference system rule base according to the reward value obtained by the global action;

[0165] Step 6: Repeat steps 2 to 5 until the result vector weights converge, so as to generate a trust assessment model according to the fuzzy inference system rule base.

[0166] In some embodiments, the calculation formula of the global action and its corresponding Q value is:

[0167]

[0168]

[0169] Among them, a(s) represents the global action, Q(s,a) represents the Q value corresponding to the global action, and w i (x) represents the trigger strength of rule i, q[i,i ′ ] represents the i-th ′ The Q value corresponding to the possible action is a i represents the action of rule i, and N represents the number of rules.

[0170] Another embodiment of the present invention provides an electronic device, such as Figure 7 As shown, the electronic device 10 includes:

[0171] One or more processors 110 and memory 120, Figure 7 A processor 110 is used as an example for description. The processor 110 and the memory 120 may be connected via a bus or other means. Figure 7 The example of connecting through bus is taken in the following.

[0172] The processor 110 is used to complete various control logics of the electronic device 10, and it can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), a single-chip microcomputer, an ARM (Acorn RICS), a graphics processing unit (GPU) or other programmable logic devices, discrete gate or transistor logic, discrete hardware components or any combination of these components. In addition, the processor 110 can also be any traditional processor, microprocessor or state machine. The processor 110 can also be implemented as a combination of computing devices, for example, a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors combined with a DSP and / or any other such configuration.

[0173] The memory 120, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer executable programs and modules, such as program instructions corresponding to the scene-adaptive permission dynamic adjustment method based on trust evaluation in the embodiment of the present invention. The processor 110 executes various functional applications and data processing of the electronic device 10 by running the non-volatile software programs, instructions and units stored in the memory 120, that is, the scene-adaptive permission dynamic adjustment method based on trust evaluation in the above method embodiment is implemented.

[0174] The memory 120 may include a program storage area and a data storage area, wherein the program storage area may store an operating platform, an application required for at least one function; the data storage area may store data created according to the use of the electronic device 10, etc. In addition, the memory 120 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 120 may optionally include a memory remotely arranged relative to the processor 110, and these remote memories may be connected to the electronic device 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0175] One or more units are stored in the memory 120, and when executed by one or more processors 110, the scene adaptive permission dynamic adjustment method based on trust evaluation in any of the above method embodiments is executed, for example, the above described Figure 1 The method comprises steps S100 to S400.

[0176] Another embodiment of the present invention provides a computer-readable storage medium, which stores computer-executable instructions. The computer-executable instructions are executed by one or more processors, for example, to execute the above-described Figure 1 The method comprises steps S100 to S400.

[0177] As an example, a computer readable storage medium can include a read-only memory (ROM), a programmable ROM (PROM), an electrically programmable ROM (EPROM), an electrically erasable ROM (EEPROM), or a flash memory. Volatile memory can include a random access memory (RAM) as an external cache memory. By way of illustration and not limitation, RAM can be obtained in many forms such as synchronous RAM (SRAM), dynamic RAM, (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synch link DRAM (SLDRAM), and direct Rambus RAM (DRRAM). The disclosed memory components or memories of the operating environment described herein are intended to include one or more of these and / or any other suitable types of memory.

[0178] In summary, the scenario-adaptive permission dynamic adjustment method, device, electronic device and storage medium based on trust evaluation provided by the present invention have the following beneficial effects:

[0179] (1) Strong adaptability to scenarios: The present invention fully considers the contextual conditions of access scenarios and can adjust permissions in a timely manner according to environmental changes to achieve more fine-grained authorization and help the system better protect sensitive information and resources. For example, in scenarios such as medical information systems or logistics management systems where patient flow and logistics flow are changeable, the present invention can fully consider contextual conditions such as patient department flow or logistics node changes to make corresponding permission adjustments;

[0180] (2) Flexible and adjustable: The present invention uses a trust assessment strategy based on fuzzy Q learning, collects real-time data for learning, and uses fuzzy Q learning methods to analyze and process data. It can select the optimal action based on the current state of the environment and the behavior of the subject. By continuously learning and updating the trust model, the strategy can adapt to different contexts and risk situations. In different scenarios, flexible trust assessments based on real-time data and environmental conditions support permission adjustments. Especially in scenarios with high data security requirements such as medical information systems or logistics management systems, the strategy can evaluate the user's trust level based on their behavior and historical data, and restrict or open their access rights accordingly;

[0181] (3) Safe and controllable: While providing dynamic permission adjustment, the present invention establishes a trust evaluation mechanism based on user historical access records. By analyzing user historical access records and behavior patterns, it can actively block illegal requests from seemingly legitimate users.

[0182] The specific implementation of the present invention described above does not constitute a limitation on the protection scope of the present invention. Any other corresponding changes and modifications made based on the technical concept of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A scenario-adaptive permission dynamic adjustment method based on trust evaluation, characterized in that: The steps include: Obtaining an access request initiated by a user, and obtaining the role authority corresponding to the user based on the access request; Obtaining a context condition corresponding to the access request, determining an access control policy based on the context condition, and adjusting the role authority corresponding to the user based on the access control policy; Based on a pre-established trust evaluation model and a trust threshold, a trust evaluation is performed on the user to obtain a trust evaluation result, wherein the trust evaluation model is established based on the user's historical behavior through a fuzzy Q learning method; Based on the adjusted role permissions corresponding to the user and the trust evaluation result, performing permission management on the access request; The trust evaluation of the user is performed based on the pre-established trust evaluation model and trust threshold to obtain a trust evaluation result, including: Based on the access request, obtaining a trust threshold corresponding to the access request; Calculate the trust evaluation score corresponding to the access request based on a pre-established trust evaluation model; Determining a trust evaluation result based on the trust evaluation score and the trust threshold; The method for establishing the trust evaluation model is: Step 1: Establish a fuzzy inference system rule base, wherein the fuzzy inference system rule base consists of N rules; Step 2: Obtain the user's historical access records to obtain the user's current status, and perform fuzzy processing on the user's current status, where the status is a five-tuple consisting of user ID, role, requested resource, permission granted, and context conditions; Step 3: Based on the current state of the user after the fuzzification process, for each triggered rule corresponding to the access request, randomly select the action of the triggered rule according to the ε-greedy strategy to calculate the action of the triggered rule and its corresponding Q value; Step 4: Based on the actions of the triggered rules and their corresponding Q values, calculate the global actions and their corresponding Q values; Step 5: executing the calculated global action and updating the current state, and after calculating the reward value, updating the result vector weight of the fuzzy inference system rule base according to the reward value obtained by the global action; Step 6: Repeat steps 2 to 5 until the result vector weights converge, so as to generate a trust assessment model according to the fuzzy inference system rule base.

2. The scenario-adaptive permission dynamic adjustment method based on trust evaluation according to claim 1 is characterized in that: The obtaining of the access request initiated by the user and obtaining the role authority corresponding to the user based on the access request includes: Obtaining an access request initiated by a user, and parsing the access request to obtain user information contained in the access request; Based on the user information, the role authority corresponding to the user is obtained.

3. The scenario-adaptive permission dynamic adjustment method based on trust evaluation according to claim 1 is characterized in that: The contextual conditions include at least time context, spatial context, user-declared context and prerequisite context. The time context is determined based on the access time of the access request, the spatial context is determined based on the location of the access request, the user-declared context is determined based on the access purpose of the access request, and the prerequisite context is determined based on the access request and a server database.

4. The scenario-adaptive permission dynamic adjustment method based on trust evaluation according to claim 3 is characterized in that: The obtaining of the context condition corresponding to the access request, determining the access control policy based on the context condition, and adjusting the role authority corresponding to the user based on the access control policy includes: Based on the access request, obtaining a context condition corresponding to the access request; Based on the context condition, obtaining an access control policy corresponding to the context condition, wherein the access policy is one or more of a rule, a specification or a logical expression; Based on the access control policy, the role permissions corresponding to the user are adjusted.

5. The scenario-adaptive permission dynamic adjustment method based on trust evaluation according to claim 1 is characterized in that: The calculation formula of the global action and its corresponding Q value is: Among them, a(s) represents the global action, Q(s, a) represents the Q value corresponding to the global action, and w i (x) represents the trigger strength of rule i, q[i, i′] represents the Q value corresponding to the i′th possible action in rule i, and a i represents the action of rule i, and N represents the number of rules.

6. A scenario-adaptive authority dynamic adjustment device based on trust evaluation, characterized in that: include: The permission acquisition module is used to obtain the access request initiated by the user and obtain the role permission corresponding to the user based on the access request; A permission adjustment module, used to obtain the context conditions corresponding to the access request, determine the access control policy based on the context conditions, and adjust the role permissions corresponding to the user based on the access control policy; A trust evaluation module, used to perform trust evaluation on the user based on a pre-established trust evaluation model and a trust threshold to obtain a trust evaluation result, wherein the trust evaluation model is established based on the user's historical behavior through a fuzzy Q learning method; A permission granting module, used to perform permission management on the access request based on the adjusted role permissions corresponding to the user and the trust assessment result; The trust assessment module is specifically used for: Based on the access request, obtaining a trust threshold corresponding to the access request; Calculate the trust evaluation score corresponding to the access request based on a pre-established trust evaluation model; Determining a trust evaluation result based on the trust evaluation score and the trust threshold; The method for establishing the trust evaluation model is: Step 1: Establish a fuzzy inference system rule base, wherein the fuzzy inference system rule base consists of N rules; Step 2: Obtain the user's historical access records to obtain the user's current status, and perform fuzzy processing on the user's current status, where the status is a five-tuple consisting of user ID, role, requested resource, permission granted, and context conditions; Step 3: Based on the current state of the user after the fuzzification process, for each triggered rule corresponding to the access request, randomly select the action of the triggered rule according to the ε-greedy strategy to calculate the action of the triggered rule and its corresponding Q value; Step 4: Based on the actions of the triggered rules and their corresponding Q values, calculate the global actions and their corresponding Q values; Step 5: executing the calculated global action and updating the current state, and after calculating the reward value, updating the result vector weight of the fuzzy inference system rule base according to the reward value obtained by the global action; Step 6: Repeat steps 2 to 5 until the result vector weights converge, so as to generate a trust assessment model according to the fuzzy inference system rule base.

7. An electronic device, characterized in that: include: Processor and memory; The memory stores a computer program executable by the processor; When the processor executes the computer program, the steps in the scenario-adaptive authority dynamic adjustment method based on trust assessment are implemented as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: include: Processor and memory; The memory stores a computer program executable by the processor; When the processor executes the computer program, the steps in the scenario-adaptive authority dynamic adjustment method based on trust assessment are implemented as described in any one of claims 1 to 5.

Images