An online learning platform data security access control method

Abstract

The present application relates to the technical field of online learning data security, and relates to an online learning platform data security access control method.The present application calls the historical operation behavior of a user before the user applies for data access after identity verification, analyzes the antecedent association of the applied access operation behavior sequence and the historical operation behavior, extracts the reference access operation behavior sequence of other users with the same attribute data in accessing the current data, compares and outputs the matching degree of the applied access operation behavior sequence, dynamically adjusts the access permission according to the analyzed risk value of the user in accessing the current data, increases the risk recognition accuracy of the user operation behavior, and according to the deviation degree of the access operation data and the user behavior portrait after the adjustment of the user access permission, determines the dynamic trust score in combination with the historical trust score of the user, realizes the quick response means when abnormal access operation occurs, and provides fine user trust management for the shuttle machine in the learning platform.

Description

Technical Field

[0001] This invention relates to the field of online learning data security technology, and specifically to a method for secure access control of data on an online learning platform. Background Technology

[0002] Online learning platforms aggregate massive amounts of course resources, user privacy information, and teaching data. Data security is the core guarantee for the stable operation of these platforms. With the increasing prevalence of mobile learning platforms, users can achieve barrier-free, immersive access across geographical and cultural boundaries, making platform access scenarios more diverse and increasing the difficulty of identifying abnormal operations, thus struggling to adapt to dynamically changing access needs. Simultaneously, user access behavior exhibits both personalized and group-wide characteristics, and a single static authorization method cannot balance data security and access convenience. Therefore, there is an urgent need for an access control method that can accurately identify user behavior and dynamically adjust permissions.

[0003] In the prior art, Chinese Patent Publication No. CN112966235A discloses a big data component access control method and system for a smart education platform. This method determines the user's authorization code through an authorization module. When a component access request is received, the request is parsed to obtain the user's target component and corresponding verification strategy. Based on the verification strategy and authorization code, the component access request is verified. If the verification is successful, the user is allowed to access the target component. Through fine-grained component access control, the security of basic components is strengthened by controlling the access process.

[0004] The existing technology has the following problems: 1. The existing technology only performs static verification through authorization codes and fixed verification strategies, and lacks correlation analysis of the user's previous operation behavior and the reference behavior of users with the same attributes. This results in inaccurate risk assessment, leading to misjudgment of normal access behavior or omission of malicious operation risk behavior. This not only affects the user access experience, but also fails to effectively prevent security risks such as data leakage and unauthorized tampering.

[0005] 2. Existing technologies do not establish user behavior profiles and dynamic trust scoring mechanisms. They rely solely on single-authorization verification to determine access permissions and cannot intervene in access control based on the trust level of users' real-time operational behavior. This results in a lack of rapid response measures when abnormal access operations occur, making it difficult to meet the real-time and flexible data security protection requirements of online learning platforms. Summary of the Invention

[0006] This invention aims to overcome the deficiencies in the prior art and provide a data security access control method for online learning platforms. Through access operation behavior analysis, dynamic permission adjustment and trust assessment, it adapts to cross-cultural access scenarios and achieves a balance between security and convenience.

[0007] The technical solution adopted by the present invention to solve its technical problem is: a data security access control method for an online learning platform, comprising: authenticating users who apply for access in the online learning platform, and obtaining the user's access application operation sequence after the authentication is successful.

[0008] Retrieve the user's historical operation behavior before requesting access to data, analyze the preceding correlation between the sequence of request access operation behavior and the historical operation behavior, extract the reference access operation behavior sequence of other users with the same attribute data when accessing the current data, and compare and output the matching degree of the request access operation behavior sequence.

[0009] Based on prior association and matching degree analysis, the risk value of a user accessing the current data is analyzed, and the user's access permissions are dynamically adjusted according to the risk value.

[0010] User behavior profiles are created based on historical access data of users accessing the same data, and the degree of deviation between access data and user behavior profiles after user access permissions are adjusted is analyzed.

[0011] A dynamic trust score is determined by combining the user's historical trust score, and corresponding access control processing is performed based on the trust level corresponding to the dynamic trust score.

[0012] Compared with the prior art, the present invention has the following beneficial effects: (1) The present invention retrieves the user's historical operation behavior before applying for access to data, analyzes the sequence of user's access operation behavior and the preceding relationship between the historical operation behavior, thereby identifying user behavior mutations and abnormal jumps, and can promptly detect deviations in user behavior after identity verification, eliminating the blind spot of data security access monitoring under static verification.

[0013] (2) This invention extracts the reference access operation behavior sequence of other users with the same attribute data when accessing the current data, compares the matching degree of the output access request operation behavior sequence, realizes the matching judgment of the access operation behavior of users with the same attribute, avoids the misjudgment problem caused by the one-size-fits-all strategy, and improves the accuracy of subsequent user access permission adjustment.

[0014] (3) Based on the prior association and matching degree analysis, the present invention analyzes the risk value of users accessing the current data, dynamically adjusts the user access permissions based on the risk value, increases the accuracy of identifying risks in user operation behavior, reduces misjudgment of normal access behavior, and improves the user data access experience and data security management of online learning platforms.

[0015] (4) Based on the establishment of user behavior profiles, this invention analyzes the degree of deviation between the access operation data after the user access permission adjustment and the user behavior profile, determines the dynamic trust score in combination with the user's historical trust score, and executes the corresponding access control processing, so as to realize a rapid response means when abnormal access operations occur, providing the shuttle with refined user trust management in the learning platform, and meeting the real-time and flexible data security protection requirements of the online learning platform. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the method steps of the present invention.

[0018] Figure 2 This is a schematic diagram illustrating the matching degree comparison steps of the application access operation behavior sequence in this invention.

[0019] Figure 3 This is a schematic diagram illustrating the steps for analyzing the deviation between access operation data and user behavior profiles after adjusting user access permissions in this invention. Detailed Implementation

[0020] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the invention. Furthermore, it should be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale.

[0021] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the invention or its application or use. Techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and apparatus should be considered part of the specification.

[0022] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0023] Please see Figure 1 As shown, the present invention provides a data security access control method for an online learning platform, including: S1, obtaining the sequence of authentication and access request operation behaviors.

[0024] Given that user accounts on online learning platforms are easily impersonated, a single verification method is insufficient to prevent unauthorized access. It is necessary to combine device and biometric verification to ensure the authenticity of users. At the same time, it is necessary to accurately capture the sequence of access behavior to provide a complete data foundation for subsequent behavior analysis.

[0025] Based on this, the specific implementation of the present invention includes: S11, verifying the identity of users applying for access to the online learning platform. The identity verification process is as follows: S111, collecting the biometric identification information and device authentication information of the users applying for access, wherein the biometric identification information includes fingerprints or facial images, which are collected by a shuttle and uploaded to the platform, and the device authentication information includes device ID, operating system version, and IP address, etc., which is collected by the platform backend without requiring manual input from the user, and verifying the consistency of the device authentication information with the platform's registered device list.

[0026] S112. After the consistency verification between the device authentication information and the registered device list on the online learning platform is passed, extract the standard biometric information pre-stored in the corresponding device in the platform.

[0027] S113. If the biometric identification information is similar to the standard biometric information pre-stored in the corresponding device, then the identity verification is deemed successful. In a specific embodiment of the present invention, a cosine similarity algorithm can be used to calculate the matching degree between the collected biometric identification information and the standard biometric information. When the similarity is greater than the similarity threshold set by the platform, the identity verification is deemed successful; otherwise, the user's biometric identification information is re-collected. If it still fails, the identity verification is deemed unsuccessful.

[0028] S12. After successful authentication, retrieve all user access requests and sort them in ascending order by timestamp to obtain the user access request sequence.

[0029] In one example, a user's access request sequence might be: (Course interaction, English course notes query). The access scenarios for course interaction and English course notes query constitute the course access scenario.

[0030] This invention reduces the risk of identity theft and effectively prevents unauthorized logins and forgery attacks by verifying the dual identity of users applying for access through both device and biometric features. It also obtains the user's access request operation sequence, providing data support for subsequent preceding association analysis and matching degree calculation, thus ensuring the legitimacy of the access subject and the integrity of the behavior data from the source.

[0031] S2, Precedence Association and Matching Degree Analysis.

[0032] Considering that user access behavior has temporal continuity and group common characteristics, analyzing the current behavior alone is prone to misjudging normal operations or missing anomalies. It is necessary to combine the previous association of historical behavior and the reference matching of users with the same attributes to improve the accuracy of behavior recognition from two dimensions.

[0033] Based on this, the present invention outputs the prior association status and matching degree through prior association analysis and group matching. The specific implementation includes: S21, retrieving the user's historical operation behavior before applying for access to data, and analyzing the prior association between the application access operation behavior sequence and the historical operation behavior, specifically as follows: S211, extracting the user's historical operation behavior records within a preset time period before applying for access to data from the learning platform database, and constructing a historical operation behavior sequence.

[0034] In one example, the historical action sequence could be: (login to platform, select course, study English course, record course notes). The login scenario is the platform access scenario, while the selection, study, and record course notes scenarios are all course access scenarios.

[0035] S212. Extract the access scenario and operation timestamp of each operation in the historical operation behavior sequence, compare them with the access scenario and operation timestamp of all operations in the request access operation behavior sequence, and determine whether the access scenario of the request access operation behavior sequence matches the historical operation behavior sequence and whether the operation behaviors are continuous.

[0036] The specific judgment process is as follows: S2121, based on the access scenario and operation timestamp of each operation in the historical operation behavior sequence, determine the representative access scenario and operation time interval range corresponding to the historical operation behavior sequence.

[0037] The representative access scenario is the access scenario that appears most frequently in the historical operation behavior sequence. For example, if the course access scenario appears most frequently in the historical operation behavior sequence, then the course access scenario will be used as the representative access scenario.

[0038] The operation time interval range is determined by obtaining the operation timestamp interval between adjacent operation behaviors in the historical operation behavior sequence, calculating the mean and standard deviation of the operation timestamp interval, and using the 3σ principle to determine the operation time interval range, covering 99% of normal operation intervals and avoiding misjudgment due to random fluctuations.

[0039] S2122. If the access scenarios of all operations in the request access operation sequence are the same as the representative access scenario, then the access scenario is determined to be a match; otherwise, the access scenario is determined to be a mismatch.

[0040] S2123. Obtain the time interval between all operations in the sequence of access request operations. If all operations are within the time interval range, they are judged as continuous operations; otherwise, they are judged as non-continuous operations.

[0041] S213. Based on the access scenario matching judgment result and the operation behavior continuity judgment result, analyze the preceding association between the sequence of access request operation behaviors and the historical operation behaviors. Specifically, if the access scenario matching judgment result is access scenario matching and the operation behavior continuity judgment result is continuous operation behavior, it is determined to be a preceding association; otherwise, it is determined to be a preceding no-association.

[0042] This invention retrieves a user's historical operational behavior before requesting access to data, analyzes the sequence of user access request operations and the preceding relationships between historical operations, thereby identifying sudden changes and abnormal jumps in user behavior. It can promptly detect deviations in user behavior after successful authentication and eliminate blind spots in data security access monitoring under static verification.

[0043] S22, and extract the reference access operation behavior sequence of other users with the same attribute data when accessing the current data, compare and output the matching degree of the access request operation behavior sequence, such as Figure 2 As shown, the comparison method is as follows: S221. Based on the user's attribute data, which includes role identity and access permission level, the user's role identity can be student, teacher, or administrator. Other users with the same attribute data are selected from the learning platform database to form a user group with the same attribute.

[0044] S222. Extract the historical operation behavior sequence of each other user in the same attribute user group when accessing the current data, filter the operation behaviors that appear in the historical operation behavior sequence of each other user, arrange them in a fixed operation order, construct the access operation behavior sequence corresponding to the same attribute user group, and use it as the reference access operation behavior sequence.

[0045] In one example, the standard sequence of operations for a teacher accessing grade management might be: querying student grades, exporting student grades, and printing student grades.

[0046] S223. Compare the request access operation behavior sequence with the reference access operation behavior sequence, filter all identical operation behaviors and their corresponding operation numbers, and calculate the matching degree of the request access operation behavior sequence.

[0047] It should be noted that the process of calculating the matching degree of the request access operation behavior sequence is as follows: First, obtain the operation order of all the same operation behavior in the request access operation behavior sequence and the reference access operation behavior sequence, and use the corresponding sequence number of the operation order as the operation sequence number.

[0048] Secondly, if the number of operations in the request access operation sequence is the same as the number of operations in the reference access operation sequence, then the operation sequence numbers of all the same operations are compared, and the proportion of the number of the same operations with the same operation sequence number is used as the matching degree.

[0049] Conversely, the number of operations in the longest sequence of the request access operation sequence and the reference access operation sequence is used as the benchmark, and the missing positions in the shortest sequence are filled with placeholders, with the placeholder being no operation.

[0050] Finally, obtain the operation sequence number of all identical operations in the completed sequence, and use the proportion of identical operations with the same operation sequence number as the matching degree.

[0051] This invention extracts reference access operation behavior sequences of other users with the same attribute when accessing the current data, compares the matching degree of the output access request operation behavior sequence, and realizes matching judgment of access operation behavior of users with the same attribute. This avoids the misjudgment problem caused by the one-size-fits-all strategy and improves the accuracy of subsequent adjustment of user access permissions.

[0052] S3. Risk value calculation and dynamic permission adjustment.

[0053] Considering that risk values ​​need to take into account both prior correlation and matching degree, a single indicator is prone to assessment bias; at the same time, permission adjustments need to be adapted to the risk level to avoid excessively restricting normal access or allowing high-risk operations. Therefore, risk values ​​need to be quantified and hierarchical permission strategies need to be formulated.

[0054] Based on this, the specific implementation of the present invention includes: First, analyzing the risk value of a user accessing the current data based on the prior association and matching degree, specifically: if the prior association between the request access operation sequence and the historical operation sequence is no prior association or the matching degree of the request access operation sequence is less than a set matching degree threshold, then the set value, such as 1, is used as the risk value; otherwise, a deviation analysis is performed on the matching degree of the request access operation sequence to determine the risk value.

[0055] Then, user access permissions are dynamically adjusted based on the risk value. Specifically, the risk value is matched with the risk value range corresponding to the preset risk level to determine the user access permissions for the corresponding risk level.

[0056] In one specific embodiment of the present invention, based on the statistical analysis of the access operation behavior sequence of all access data in the learning platform database, the proportion of all access data with a matching degree greater than 80% is 99%, so the matching degree threshold is set to 80%. The implementer can also adjust the matching degree threshold himself.

[0057] The risk value is determined as follows: the difference between the matching degree of the request access operation behavior sequence and the matching degree corresponding to the complete match (default is 100%) is obtained, and the product of the difference and the maximum risk value is used as the risk value. For example, when the maximum risk value is a set value of 1, the risk value is the difference between the matching degree and the matching degree corresponding to the complete match.

[0058] In one example, the user access permissions for a preset risk level are as follows: when the risk level is low (0 ≤ risk value ≤ 0.2), if the user requesting access does not have permission to access the current data, then the user is granted permission to access the current data; otherwise, the user access permissions are retained.

[0059] When the risk level is medium risk (0.2 < risk value ≤ 0.5), the user's access to the current data is restricted.

[0060] When the risk level is high (0.5 < risk value ≤ 1), all permissions will be temporarily frozen, and secondary identity verification is required.

[0061] This invention analyzes the risk value of a user accessing current data based on prior association and matching degree, and dynamically adjusts user access permissions based on the risk value. This increases the accuracy of identifying risks in user operation behavior, reduces misjudgments of normal access behavior, and improves the user data access experience and data security management of online learning platforms.

[0062] S4. User behavior profile establishment and deviation analysis.

[0063] Considering that user access behavior has personalized temporal characteristics, historical data of a single dimension cannot build an accurate profile, and the degree of deviation needs to be judged by quantitative indicators to avoid subjective misjudgment. It is necessary to subdivide features by time period, standardize and vectorize them, and clarify the deviation threshold of each dimension to provide an accurate basis for dynamic trust scoring.

[0064] Based on this, the present invention achieves abnormal behavior identification through multi-dimensional feature modeling and deviation quantification. The specific implementation includes: S41, establishing a user behavior profile based on the user's historical access operation data of accessing the same data.

[0065] The steps for establishing the user behavior profile are as follows: First, retrieve the historical records of each user accessing the same data from the learning platform database, and classify each historical record by time period.

[0066] It should be noted that the time period classification method in this embodiment of the invention is as follows: weekdays are divided into morning period (8:00-12:00), noon period (12:00-14:00), and evening period (14:00-22:00), and rest days are divided into morning period (9:00-12:00), afternoon period (12:00-18:00), and night period (18:00-23:00). Implementers may also classify the time period themselves.

[0067] Then, the access duration, permission call combination, and operation characteristics of each historical access operation data in different time periods are obtained to form user access operation data vectors for different time periods, establish user behavior profiles, and regularly update user behavior profiles based on the latest user access behavior data.

[0068] In one specific embodiment of the present invention, the access duration is calculated by averaging the access durations of all historical records within different time periods.

[0069] Permission call combinations: Statistically analyze the permission call combinations of all historical records within different time periods, such as "Note query and download", "Note query, collection and modification", and "Note query and sharing". Calculate the frequency of occurrence of each permission call combination and take the permission call combination with the highest frequency as the permission call combination for the corresponding time period.

[0070] Operational characteristics: Statistically analyze the operational characteristics of different time period permission call combinations in each historical record, such as note query duration and frequency, number of note downloads, number of note favorites, etc. Filter all operational characteristics ranges and use them as the operational characteristics for the corresponding time period.

[0071] S42. Analyze the degree of deviation between access operation data and user behavior profiles after user access permission adjustments, such as... Figure 3 As shown, the specific analysis method is as follows: S421, extract the access time period, access duration, permission call combination and operation link features from the access operation data after the user access permission is adjusted, and construct the current access feature vector.

[0072] S422. Compare the current access feature vector with the user access operation data vector of the corresponding time period in the user behavior profile, and calculate the deviation of each dimension feature.

[0073] The access duration deviation is calculated as follows: obtain the absolute difference in access duration between the current access feature vector and the user access operation data vector, and use the ratio of the absolute difference in access duration to the access duration of the user access operation data vector as the access duration deviation.

[0074] The deviation of permission call combination is calculated as follows: if the permission call combination in the current access feature vector belongs to the permission call combination in the user access operation data vector, then the deviation of permission call combination is recorded as 0; otherwise, the deviation of permission call combination is recorded as 1.

[0075] The operation link feature deviation is calculated as follows: the number of features in the current access feature vector that are not within the corresponding operation link feature range is selected, and the ratio of this number to the total number of operation link features in the current access feature vector is taken as the operation link feature deviation.

[0076] In one specific embodiment of the present invention, for example, the allowable deviation for access duration can be set to ±10% of the access duration of the user access operation data vector; the allowable deviation for permission call combination is 0, i.e., no deviation; and the allowable deviation for operation link features can be set to ±10%. In other embodiments, the implementer may also set the corresponding allowable deviations as needed.

[0077] S423. If the deviation of any dimension feature exceeds the corresponding allowable deviation, then the degree of deviation between the access operation data and the user behavior profile is considered a behavioral deviation; otherwise, it is considered unbiased.

[0078] S5. Dynamic trust scoring and access control processing.

[0079] Considering that user trust levels need to take into account both long-term behavioral performance and real-time operational deviations, a single historical score cannot reflect dynamic risks. At the same time, trust levels need to be handled with differentiated strategies to ensure rapid response to abnormal behavior and avoid data security risks.

[0080] Based on this, the present invention achieves precise access control through dynamic adjustment and hierarchical processing of scores. The specific implementation includes: determining a dynamic trust score by combining the user's historical trust score, and performing corresponding access control processing based on the trust level corresponding to the dynamic trust score.

[0081] It should be noted that the dynamic trust score determination process is as follows: S51, when there is no bias, the user's historical trust score is used as the dynamic trust score; when there is behavioral bias, the dynamic trust score is obtained by attenuation calculation based on the number of bias dimension features. For example, when the number of bias dimension features is 1, the user's historical trust score is reduced by 1, and the maximum number of bias dimension features is 3.

[0082] S52. Compare and match the dynamic trust score with the preset trust level score range to determine the user's current trust level.

[0083] S53. If a user's current trust level is lower than the trust level corresponding to the user's historical trust score, then revoke the user's access permissions and lock the user's operation behavior, and send a risk warning notification to the learning platform.

[0084] In one embodiment of the present invention, the preset trust level can be set to a high trust level, a medium trust level, and a low trust level, and the scoring range of the preset trust level is set as follows: High trust level: 80 points ≤ dynamic trust score ≤ 100 points; Medium trust level: 60 points ≤ dynamic trust score < 80 points; Low trust level: 0 points ≤ dynamic trust score < 60 points.

[0085] In other embodiments, the implementer can adjust the scoring range of the preset trust level as needed.

[0086] This invention establishes user behavior profiles, analyzes the degree of deviation between access operation data and user behavior profiles after user access permissions are adjusted, determines dynamic trust scores by combining user historical trust scores, and executes corresponding access control processing. This enables rapid response measures when abnormal access operations occur, providing the shuttle with refined user trust management in the learning platform, and meeting the real-time and flexible data security protection requirements of the online learning platform.

[0087] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0088] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0089] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0090] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0091] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A data security access control method for an online learning platform, characterized in that, include: Verify the identity of users who apply for access to the online learning platform, and obtain the user's access request operation sequence after the identity verification is successful; Retrieve the user's historical operation behavior before requesting access to data, analyze the preceding correlation between the request access operation behavior sequence and the historical operation behavior, extract the reference access operation behavior sequence of other users with the same attribute data when accessing the current data, and compare and output the matching degree of the request access operation behavior sequence. Based on prior association and matching degree analysis, the risk value of users accessing current data is analyzed, and user access permissions are dynamically adjusted according to the risk value. User behavior profiles are created based on historical access data of users accessing the same data, and the degree of deviation between access data and user behavior profiles after user access permissions are adjusted is analyzed. A dynamic trust score is determined by combining the user's historical trust score, and corresponding access control processing is performed based on the trust level corresponding to the dynamic trust score. The analysis of the sequence of access request actions and their preceding historical actions is as follows: Extract historical operation records of users within a preset time period before requesting access to data from the learning platform database, and construct a historical operation sequence. Extract the access scenario and operation timestamp of each operation in the historical operation sequence, compare them with the access scenario and operation timestamp of all operations in the request access operation sequence, and determine whether the access scenario of the request access operation sequence matches the historical operation sequence and whether the operation is continuous. Based on the access scenario matching judgment results and the operation behavior continuity judgment results, we analyze the preceding association between the sequence of access request operation behavior and the historical operation behavior. The matching degree comparison method for the access request operation behavior sequence is as follows: Based on the user's attribute data, including role identity and access permission level, other users with the same attribute data are selected from the learning platform database to form a user group with the same attribute. Extract the historical operation behavior sequence of other users in the same attribute user group when accessing the current data, construct the access operation behavior sequence corresponding to the same attribute user group, and use it as the reference access operation behavior sequence. The sequence of request access operation behavior is compared with the sequence of reference access operation behavior, all identical operation behaviors and their corresponding operation numbers are filtered out, and the matching degree of the request access operation behavior sequence is calculated. The steps for establishing a user behavior profile are as follows: Retrieve historical records of users accessing the same data from the learning platform's database and categorize these historical records by time period. By acquiring the access duration, permission call combinations, and operation characteristics from historical access operation data of each historical record within different time periods, user access operation data vectors for different time periods are constructed to establish user behavior profiles.

2. The data security access control method for an online learning platform according to claim 1, characterized in that: The authentication process is as follows: Collect the biometric identification information and device authentication information of users who apply for access, and verify the consistency of the device authentication information with the list of registered devices on the platform; Once the consistency verification between the device authentication information and the registered device list on the online learning platform is passed, the standard biometric information pre-stored in the corresponding device in the platform is extracted. If the biometric identification information is similar to the standard biometric information pre-stored in the corresponding device, the identity verification is deemed successful.

3. The data security access control method for an online learning platform according to claim 1, characterized in that: The process for judging the access scenario matching result and the continuous judgment result of the operation behavior is as follows: Based on the access scenario and operation timestamp of each operation in the historical operation sequence, determine the representative access scenario and operation time interval range corresponding to the historical operation sequence. If the access scenarios of all operations in the request access operation sequence are the same as the representative access scenario, then the access scenario is considered to match; otherwise, the access scenario is considered to be mismatched. Get the time interval between all operations in the request access operation sequence. If they are all within the time interval range, they are judged as continuous operations; otherwise, they are judged as non-continuous operations.

4. The data security access control method for an online learning platform according to claim 1, characterized in that: The process of calculating the matching degree of the request access operation behavior sequence is as follows: Obtain the operation order of all identical operation behaviors in the request access operation behavior sequence and the reference access operation behavior sequence, and use the corresponding sequence number of their operation order as the operation sequence number; If the number of operations in the request access operation sequence is the same as the number of operations in the reference access operation sequence, then the operation sequence numbers of all the same operations are compared, and the proportion of the number of the same operations with the same operation sequence numbers is used as the matching degree. Conversely, the number of operations in the longest sequence of the request access operation sequence and the reference access operation sequence is used as the benchmark, and placeholders are used to fill in the missing positions of operations in the shortest sequence. Obtain the operation sequence number of all identical operations in the completed sequence, and use the proportion of identical operations with the same operation sequence number as the matching degree.

5. The data security access control method for an online learning platform according to claim 1, characterized in that: The method for dynamically adjusting user access permissions is as follows: If the sequence of access request actions has no preceding association with the previous actions, or if the matching degree of the access request action sequence is less than the set matching degree threshold, then the set value will be used as the risk value. Conversely, a deviation analysis is performed on the matching degree of the access request operation sequence to determine the risk value; The risk value is matched with the risk value range corresponding to the preset risk level to determine the user access permissions for the corresponding risk level.

6. The data security access control method for an online learning platform according to claim 1, characterized in that: The method for analyzing the deviation between the access operation data and the user behavior profile is as follows: Extract access time period, access duration, permission call combination and operation steps from the access operation data after the user access permission is adjusted, and construct the current access feature vector; The current access feature vector is compared with the user access operation data vector of the corresponding time period in the user behavior profile, and the deviation of each dimension feature is calculated. If the deviation of any dimension feature exceeds the corresponding allowable deviation, then the degree of deviation between the access operation data and the user behavior profile is considered behavioral deviation; otherwise, it is considered unbiased.

7. The data security access control method for an online learning platform according to claim 6, characterized in that: The dynamic trust scoring determination process is as follows: When there is no bias, the user's historical trust score is used as the dynamic trust score. When there is a behavioral bias, the dynamic trust score is obtained by decay calculation based on the number of features of the bias dimension. The dynamic trust score is compared and matched with the preset trust level score range to determine the user's current trust level. If a user's current trust level is lower than the trust level corresponding to their historical trust score, then the user's access permissions will be revoked and the user's actions will be locked. At the same time, a risk warning notification will be sent to the learning platform.

Images