Privacy access control method and system in Internet of Things environment

By combining data desensitization and privacy access control technology in the IoT environment, using the IoT session desensitization network trained with significant feature errors, the problem that privacy access control in traditional methods is difficult to cover complex scenarios and process pressures is achieved. Efficient and secure privacy access control improves the security of IoT systems and data processing intelligence.

CN119094152AActive Publication Date: 2024-12-06CHENGDU LECHAOREN TECH CO LTD

Patent Information

Application Number
CN202410984443.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-12-06
Estimated Expiration
2044-07-22

AI Technical Summary

Technical Problem

In traditional IoT environments, privacy access control methods are difficult to effectively cover all situations in complex and changeable scenarios, resulting in the risk of privacy leakage and face huge processing pressure, which is difficult to meet the security needs of massive data.

Method used

By combining data desensitization and privacy access control technology, the IoT session log is obtained and loaded into the IoT session desensitization network based on significance feature error training, data desensitization is performed, and the target session log optimization is used after desensitization Privacy access control network to achieve efficient and secure privacy access control.

Benefits of technology

Effectively remove or replace sensitive information in the log, ensure that the desensitized data sequence matches the original data, improve privacy protection capabilities and intelligence of data processing, significantly improve the accuracy and efficiency of privacy access control, and enhance the object. Security of networked systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119094152A_ABST
    Figure CN119094152A_ABST
Patent Text Reader

Abstract

The invention provides a privacy access control method and system in an Internet of Things environment, and the method comprises the steps: firstly obtaining an Internet of Things session log of a first Internet of Things service environment, and loading the Internet of Things session log into an Internet of Things session desensitization network obtained based on significance feature error training; and the Internet of Things session desensitization network ensures that the salient feature data in the desensitized target session log is matched with the original log. And then, optimizing the privacy access control network of the second Internet of Things service environment by using the desensitized target session logs, so as to improve the accuracy and efficiency of privacy access control. Therefore, the data desensitization technology and the privacy access control technology are combined, the key data features are reserved while the privacy of the user is protected, and therefore safe and efficient privacy access control under the environment of the Internet of Things is achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of Internet of Things, and more specifically, to a privacy access control method and system in an Internet of Things environment. Background Art

[0002] With the rapid development of IoT technology, more and more devices are connected to the Internet, forming a huge IoT system. The communication and interaction between these devices generate a large amount of data, including the privacy information of many users. How to protect this privacy information from being leaked while effectively processing data and controlling access has become an urgent problem to be solved in the development of IoT technology.

[0003] In the traditional IoT service environment, privacy access control usually relies on preset rules and policies. However, this approach often seems powerless when faced with complex and ever-changing IoT environments and massive amounts of data. On the one hand, preset rules and policies are difficult to cover all scenarios and situations, which easily leads to the risk of privacy leakage; on the other hand, with the continuous increase of IoT devices and the continuous growth of data volume, traditional privacy access control methods are facing huge processing pressure. Summary of the invention

[0004] In view of this, the purpose of this application is to provide a privacy access control method and system in an Internet of Things environment, aiming to achieve efficient and secure privacy access control in an Internet of Things environment by combining data desensitization and privacy access control technology.

[0005] According to a first aspect of the present application, a privacy access control method in an Internet of Things environment is provided, the method comprising: Obtain the IoT session log of the first IoT service environment; Loading the IoT session log into an IoT session desensitization network, wherein the IoT session desensitization network is generated by network learning based on significant feature errors, wherein the significant feature errors are determined based on sample significant feature data in a first desensitized data sequence generated after data desensitization of a first sample IoT session data sequence in a first IoT service environment, and sample significant feature data in a second desensitized data sequence generated after data desensitization of a second sample IoT session data sequence in a second IoT service environment; Obtaining a target session log of the second Internet of Things service environment generated by the Internet of Things session desensitization network performing data desensitization on the Internet of Things session log, wherein the significant feature data included in the target session log matches the significant feature data included in the Internet of Things session log; After optimizing the privacy access control network of the second Internet of Things service environment based on the target session log of the second Internet of Things service environment, the target privacy access control task is performed based on the optimized privacy access control network.

[0006] In a possible implementation manner of the first aspect, before loading the Internet of Things session log into the Internet of Things session desensitization network, the method further includes: Acquire the first sample IoT session data sequence, the second sample IoT session data sequence, and a candidate neural network, wherein the candidate neural network includes a first desensitization prediction unit and a second desensitization prediction unit, and the sample IoT session data in the first sample IoT session data sequence and the second sample IoT session data sequence are both annotated with prior label knowledge for indicating the sample significant feature data; The first sample IoT session data sequence is loaded into the first desensitization prediction unit, and data desensitization is performed to generate a first desensitized data sequence of the second IoT service environment annotated with the sample significant feature data, and the second sample IoT session data sequence is loaded into the second desensitization prediction unit, and data desensitization is performed to generate a second desensitized data sequence of the first IoT service environment annotated with the sample significant feature data; Determining the significant feature error based on the sample significant feature data in the second desensitized data sequence and the sample significant feature data in the first desensitized data sequence; The candidate neural network is trained based on the significant feature error to generate the Internet of Things session desensitization network.

[0007] In a possible implementation of the first aspect, the acquiring the first sample IoT session data sequence, the second sample IoT session data sequence, and the candidate neural network includes: Obtaining sample IoT session data of the first IoT service environment, and obtaining sample IoT session data of the second IoT service environment; Performing prior knowledge annotation of the sample significant feature data on the acquired sample IoT session data of the first IoT service environment to generate the first sample IoT session data sequence; Prior knowledge annotation of the sample significant feature data is performed on part of the sample Internet of Things session data obtained in the second Internet of Things service environment to generate the second sample Internet of Things session data sequence.

[0008] In a possible implementation of the first aspect, determining the significant feature error based on the sample significant feature data in the second desensitized data sequence and the sample significant feature data in the first desensitized data sequence includes: Generate a first significant feature error based on a feature distance between sample significant feature data in the first desensitized data sequence and sample significant feature data in the first sample IoT session data sequence; Generate a second significant feature error based on a feature distance between sample significant feature data in the second desensitized data sequence and sample significant feature data in the second sample IoT session data sequence; The significant feature error is generated based on the first significant feature error and the second significant feature error.

[0009] In a possible implementation manner of the first aspect, before training the candidate neural network based on the significant feature error, the method further includes: Loading the first desensitized data sequence into the second desensitization prediction unit to perform data desensitization prediction to generate a first reconstructed session data sequence for the first Internet of Things service environment; Loading the second desensitized data sequence into the first desensitization prediction unit to perform data desensitization prediction, and generating a second reconstructed session data sequence for the second Internet of Things service environment; Generate a first reconstruction error parameter based on the correlation between the first reconstructed session data sequence and the first sample IoT session data sequence; generating a second reconstruction error parameter based on the correlation between the second reconstructed session data sequence and the second sample IoT session data sequence; generating the reconstruction error parameter based on the first reconstruction error parameter and the second reconstruction error parameter; The training of the candidate neural network based on the significant feature error comprises: A global error parameter corresponding to the candidate neural network is generated based on the reconstruction error parameter and the significant feature error, and the candidate neural network is trained based on the global error parameter.

[0010] In a possible implementation of the first aspect, the candidate neural network further includes a first identification unit for identifying the validity of the session log on the second Internet of Things service environment, and a second identification unit for identifying the validity of the session log on the first Internet of Things service environment; Before training the candidate neural network based on the global error parameter, the method further includes: Loading the first desensitized data sequence into the first identification unit to generate a first identification result; Loading the second desensitized data sequence into the second identification unit to generate a second identification result; Loading the second sample IoT session data sequence into the first authentication unit to generate a third authentication result; Loading the first sample IoT session data sequence into the second authentication unit to generate a fourth authentication result; generating a first authentication comparison error parameter based on the first authentication result and the third authentication result; generating a second authentication comparison error parameter based on the second authentication result and the fourth authentication result; generating an identification comparison error parameter based on the first identification comparison error parameter and the second identification comparison error parameter; In a possible implementation of the first aspect, generating a global error parameter corresponding to the candidate neural network based on the reconstruction error parameter and the significant feature error includes: Obtaining a training participation coefficient of the reconstruction error parameter and a training participation coefficient of the significant feature error; Based on the training participation coefficient of the reconstruction error parameter and the training participation coefficient of the significant feature error, fusing the reconstruction error parameter and the significant feature error to generate a fusion error parameter; A global error parameter corresponding to the candidate neural network is generated based on the fusion error parameter and the discrimination comparison error parameter.

[0011] In a possible implementation of the first aspect, the training the candidate neural network based on the global error parameter includes: Locking the weight configuration information of the first identification unit and the second identification unit, and training the first desensitization prediction unit and the second desensitization prediction unit based on the global error parameter; The weight configuration information of the first desensitization prediction unit and the second desensitization prediction unit is locked, and the first identification unit and the second identification unit are trained based on the identification comparison error parameter.

[0012] In a possible implementation manner of the first aspect, obtaining the target session log of the second Internet of Things service environment generated by performing data desensitization on the Internet of Things session log by the Internet of Things session desensitization network includes: Obtain a target session log of the second Internet of Things service environment generated by data desensitization performed by the first desensitization prediction unit of the Internet of Things session desensitization network.

[0013] In a possible implementation of the first aspect, the significant feature data included in the target session log is a feature part marked for the privacy access control task; and the step of optimizing the privacy access control network of the second Internet of Things service environment based on the target session log of the second Internet of Things service environment includes: Acquire a privacy access control network of a second Internet of Things service environment, wherein the privacy access control network is used to perform the privacy access control task according to the significant feature data; Acquire a sample learning data sequence, wherein the sample learning data sequence includes a target session log and annotation data indicating the significant feature data; The privacy access control network is optimized based on the sample learning data sequence.

[0014] According to the second aspect of the present application, a privacy access control system in an Internet of Things environment is provided, wherein the privacy access control system in the Internet of Things environment comprises a machine-readable storage medium and a processor, wherein the machine-readable storage medium stores machine-executable instructions, and when the processor executes the machine-executable instructions, the privacy access control system in the Internet of Things environment implements the aforementioned privacy access control method in the Internet of Things environment.

[0015] According to a third aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed, the aforementioned privacy access control method in the Internet of Things environment is implemented.

[0016] According to any one of the above aspects, the technical effect of the present application is: By obtaining the IoT session log of the first IoT service environment and loading the IoT session log into the IoT session desensitization network obtained by training based on significant feature errors, the data can be effectively desensitized, which can not only remove or replace sensitive information in the log, but also ensure that the significant feature data in the desensitized data sequence matches the original data, thereby protecting the user's privacy while retaining the key features of the data. Furthermore, the privacy access control network of the second IoT service environment is optimized using the desensitized target session log, which can significantly improve the accuracy and efficiency of privacy access control. The optimized privacy access control network can more accurately identify abnormal behaviors in the IoT environment and adjust the access control strategy in time, thereby improving the security and privacy protection capabilities of the entire IoT system. Therefore, by combining data desensitization and privacy access control network, the security of the IoT service environment and the intelligence of data processing are effectively improved while protecting user privacy. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 A flowchart of a privacy access control method in an Internet of Things environment provided in an embodiment of the present application.

[0019] Figure 2 This is a schematic diagram of the component structure of the privacy access control system in the Internet of Things environment provided by an embodiment of the present application. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical scheme and advantages of the embodiment of the present application clearer, the technical scheme in the embodiment of the present application will be clearly and completely described below according to the drawings in the embodiment of the present application. It should be understood that the drawings in the present application only serve the purpose of explanation and description, and are not used to limit the scope of protection of the present application. In addition, it should be understood that the schematic drawings are not drawn in real proportion. The flowchart used in this application shows the operations implemented according to some embodiments of the embodiment of the present application. It should be understood that the operations of the flowchart can be implemented out of sequence, and the steps without logical context can be reversed in order or implemented simultaneously. In addition, those skilled in the art can add multiple other operations to the flowchart under the guidance of the content of the present application, and can also destroy multiple operations from the flowchart.

[0021] Before introducing the embodiments of the present application in detail, the meanings of some terms involved in the following embodiments are first explained.

[0022] The following is a detailed explanation and example of the first IoT service environment, IoT session log, sample significant feature data, and significant feature data: 1. First IoT service environment, second IoT service environment The first IoT service environment and the second IoT service environment refer to a specific IoT application scenario, which includes multiple IoT devices that are interconnected through the network and can transmit and interact with each other. The first IoT service environment and the second IoT service environment may be a smart home system, an industrial IoT monitoring system, a smart city traffic management system, etc. In this environment, various devices such as sensors, actuators, controllers, etc. are intelligently managed and controlled through IoT technology.

[0023] For example, suppose a smart home system is the first IoT service environment. In this smart home system, there are smart light bulbs, smart thermostats, smart door locks, smart security cameras and other devices. These devices can be connected through wireless networks to achieve remote control and data interaction. Users can control these devices through mobile phone apps or smart voice assistants, such as turning on lights, adjusting temperatures, and viewing real-time images from security cameras.

[0024] 2. IoT Session Logs IoT session logs refer to data records generated when various devices communicate and interact with each other in the IoT service environment. These logs contain key information such as connection requests, data transmission, and control instructions between devices, and are an important basis for analyzing IoT system behavior, optimizing system performance, and ensuring system security.

[0025] For example, in a smart home system, IoT session logs may record the following: when a smart light bulb is turned on or off, how a smart thermostat adjusts the indoor temperature, abnormal events captured by a smart security camera, etc. These log data are usually recorded in the form of timestamp, device ID, event type, data content, etc., to facilitate subsequent analysis and processing.

[0026] 3. Sample salient feature data For example, sample significant feature data refers to those data features in IoT session data that are key to analyzing or predicting the behavior of IoT systems. These features may include specific communication patterns, data transmission peaks, device status changes, etc. They can reflect the operating status of IoT systems, user behavior habits, and potential security risks.

[0027] For example, in a smart home system, sample significant feature data may include: the switching frequency of smart light bulbs in a specific time period (reflecting the user's living habits), the temperature adjustment mode of smart thermostats in extreme weather (reflecting the system's adaptive ability), and abnormal moving objects captured by smart security cameras at night (reflecting potential security risks), etc. These data features are of great significance for understanding the behavior patterns of IoT systems, optimizing system performance, and preventing security risks.

[0028] 4. Significant feature data Significant feature data refers to data features in IoT session logs that are identified after analysis and processing and are critical for understanding and predicting the behavior of IoT systems. These data features are not only representative in the current dataset, but can also be generalized to a wider range of IoT service environments. The difference between significant feature data and sample significant feature data is that the former is conclusive data obtained after analysis and processing, while the latter is part of the original data.

[0029] For example, in a smart home system, through the analysis and processing of IoT session logs, the following significant feature data may be identified: the switching frequency of smart light bulbs at night is significantly higher than that during the day (reflecting the user's work and rest habits), the energy consumption of smart thermostats in winter is significantly higher than that in summer (reflecting the impact of seasonal changes on energy consumption), and the number of abnormal events captured by smart security cameras on weekends has increased significantly (reflecting changes in user activity patterns on weekends). These data features are instructive for optimizing the performance of smart home systems, improving user experience, and preventing security risks.

[0030] Figure 1 The flowchart of the privacy access control method and system in the Internet of Things environment provided by the embodiment of the present application is shown. It should be understood that in other embodiments, the order of some steps in the privacy access control method in the Internet of Things environment of this embodiment can be shared with each other according to actual needs, or some steps can be omitted or maintained. The detailed steps of the privacy access control method in the Internet of Things environment include: In step S110, an IoT session log of a first IoT service environment is obtained.

[0031] In this embodiment, the privacy access control system in the IoT environment acts as a server to collect IoT session logs from a first IoT service environment (e.g., a smart home system). These IoT session logs record the communication and interaction data between various IoT devices (e.g., smart light bulbs, smart thermostats, smart security cameras, etc.) in the first IoT service environment. These data may include information such as connection requests, data transmission, and control instructions between devices. The server extracts these IoT session logs from the IoT service environment through a secure communication protocol and prepares them for subsequent processing.

[0032] Step S120: Load the IoT session log into an IoT session desensitizing network, wherein the IoT session desensitizing network is generated by network learning based on significant feature errors, wherein the significant feature errors are determined based on sample significant feature data in a first desensitized data sequence generated after data desensitization of a first sample IoT session data sequence in a first IoT service environment, and sample significant feature data in a second desensitized data sequence generated after data desensitization of a second sample IoT session data sequence in a second IoT service environment.

[0033] In this embodiment, the server loads the collected IoT session logs into a specially trained IoT session desensitization network, which is generated by network learning based on significant feature errors. During the training process, the IoT session desensitization network uses a sample IoT session data sequence (desensitized) from the first IoT service environment and a sample IoT session data sequence from the second IoT service environment (which may be another smart home system or industrial IoT environment). The sample significant feature data in these two data sequences are used to determine the significant feature errors, thereby guiding the training of the IoT session desensitization network. After loading the logs into the IoT session desensitization network, the IoT session desensitization network will desensitize these logs to remove or replace the sensitive information therein.

[0034] Step S130, obtaining a target session log of the second Internet of Things service environment generated by the Internet of Things session desensitizing network performing data desensitization on the Internet of Things session log, wherein the significant feature data included in the target session log matches the significant feature data included in the Internet of Things session log.

[0035] After being processed by the IoT session desensitizing network, the server obtains the desensitized target session logs, which have now been converted into a format suitable for the second IoT service environment, and the significant feature data therein matches the significant feature data in the original IoT session logs, which means that key features such as important interaction information and device behavior patterns have been retained during the desensitization process, but specific sensitive data (such as user identity information, device serial number, etc.) has been replaced or deleted.

[0036] Step S140: After optimizing the privacy access control network of the second Internet of Things service environment based on the target session log of the second Internet of Things service environment, the target privacy access control task is executed based on the optimized privacy access control network.

[0037] In this embodiment, the server now uses these desensitized target session logs to optimize the privacy access control network in the second IoT service environment, which is responsible for limiting or allowing access to IoT devices and data according to the set rules. By analyzing the significant feature data in the target session logs, the privacy access control network can learn which device interactions are common and which are unusual in the second IoT service environment, and update its access control policies accordingly. For example, if a device is found to frequently send large amounts of data late at night, this may be an abnormal behavior, and the privacy access control network can adjust its policy accordingly to limit access rights for such behavior. The optimized privacy access control network will more effectively protect data security and user privacy in the second IoT service environment.

[0038] Based on the above steps, by obtaining the IoT session log of the first IoT service environment and loading the IoT session log into the IoT session desensitization network obtained by training based on significant feature errors, the data can be effectively desensitized, which can not only remove or replace sensitive information in the log, but also ensure that the significant feature data in the desensitized data sequence matches the original data, thereby protecting the user's privacy while retaining the key features of the data. Further, the privacy access control network of the second IoT service environment is optimized using the desensitized target session log, which can significantly improve the accuracy and efficiency of privacy access control. The optimized privacy access control network can more accurately identify abnormal behaviors in the IoT environment and adjust the access control strategy in time, thereby improving the security and privacy protection capabilities of the entire IoT system. Therefore, by combining data desensitization and privacy access control networks, the security of the IoT service environment and the intelligence of data processing are effectively improved while protecting user privacy.

[0039] In a possible implementation, before loading the IoT session log into the IoT session desensitization network, the method further includes: Step S101: Obtain the first sample IoT session data sequence, the second sample IoT session data sequence, and a candidate neural network, wherein the candidate neural network includes a first desensitization prediction unit and a second desensitization prediction unit, and the sample IoT session data in the first sample IoT session data sequence and the second sample IoT session data sequence are both annotated with prior label knowledge for indicating the sample significant feature data.

[0040] In this embodiment, the server first accesses its data storage and retrieves sample IoT session data sequences of the first IoT service environment and the second IoT service environment, which are previously collected and labeled and include communication sessions between various IoT devices. The sample IoT session data in each sample IoT session data sequence are labeled with prior label knowledge, which indicates sample salient features in the data, such as important communication modes, data transmission types, or key device behaviors.

[0041] At the same time, the server also loads a candidate neural network, which is designed for the desensitization task of IoT session data and contains two main desensitization prediction units: the first desensitization prediction unit and the second desensitization prediction unit, which are used to process session data from different IoT service environments respectively.

[0042] Step S102: Load the first sample IoT session data sequence into the first desensitization prediction unit, perform data desensitization to generate a first desensitized data sequence of the second IoT service environment annotated with the sample significant feature data, and load the second sample IoT session data sequence into the second desensitization prediction unit, perform data desensitization to generate a second desensitized data sequence of the first IoT service environment annotated with the sample significant feature data.

[0043] In this embodiment, the server starts to load the first example IoT session data sequence into the first desensitization prediction unit, and after the first example IoT session data sequence is desensitized, the first desensitized data sequence of the second IoT service environment is generated. In this process, the desensitization prediction unit will identify and retain the significant features in the data, while removing or replacing other sensitive information.

[0044] Similarly, the server also loads the second sample IoT session data sequence into the second desensitizing prediction unit, performs data desensitization processing, and generates a second desensitized data sequence of the first IoT service environment.

[0045] Step S103: determining the significant feature error based on the sample significant feature data in the second desensitized data sequence and the sample significant feature data in the first desensitized data sequence.

[0046] In this embodiment, the server now generates two sets of desensitized data sequences: a first desensitized data sequence and a second desensitized data sequence. In order to evaluate the effect of the desensitization process and further optimize the desensitization network, the server can compare the sample significant feature data in the first desensitized data sequence and the second desensitized data sequence.

[0047] By calculating the difference between the sample significant feature data in the second desensitized data sequence and the sample significant feature data in the first desensitized data sequence, the server determines the significant feature error, which reflects the degree of retention of significant features during the desensitization process. The smaller the error, the stronger the ability of the desensitization network to retain significant features when processing data.

[0048] Step S104: training the candidate neural network based on the significant feature error to generate the IoT session desensitization network.

[0049] After determining the significant feature error as an evaluation metric, the server begins to use this significant feature error to train the candidate neural network. Through the back-propagation algorithm, the server adjusts the weights and parameters in the candidate neural network to optimize with the goal of minimizing the significant feature error.

[0050] During the training process, the server can iterate the entire dataset multiple times, and each iteration will adjust the parameters of the candidate neural network based on the current significant feature error. As the training progresses, the candidate neural network gradually learns how to better preserve the significant features of the data during the desensitization process.

[0051] Finally, when the training reaches the preset stopping conditions (such as the number of iterations, error convergence, etc.), the server will obtain an optimized IoT session desensitization network, which can protect user privacy while retaining the significant features in the IoT session data to the maximum extent, providing strong support for subsequent data analysis and privacy access control.

[0052] In a possible implementation, step S101 may include: Step S1011, obtaining sample IoT session data of the first IoT service environment, and obtaining sample IoT session data of the second IoT service environment.

[0053] In this embodiment, the server collects a series of sample IoT session data from the first IoT service environment through a secure network connection. The sample IoT session data may include various sensor readings, communication records between devices, user-device interaction logs, etc. The server ensures that encryption and verification mechanisms are used during data transmission to ensure data integrity and security.

[0054] Similar to the first IoT service environment, the server also obtains another series of sample IoT session data from the second IoT service environment in a secure manner. These sample IoT session data also contain rich IoT activity information, but may be different from the data of the first IoT service environment because the two environments may use different devices, communication protocols or data processing logic.

[0055] Step S1012: annotating the sample IoT session data of the first IoT service environment with prior knowledge of the sample significant feature data to generate the first sample IoT session data sequence.

[0056] In this embodiment, the server uses a built-in or externally provided expert system to perform a detailed analysis of the sample IoT session data collected from the first IoT service environment. By analyzing the characteristics of the data, such as communication frequency, data transmission volume, device behavior patterns, etc., the server identifies significant features in the data, such as abnormal communication, key device status changes, etc. Then, the server marks these significant features on the corresponding data in the form of prior knowledge labels, thereby generating the first sample IoT session data sequence.

[0057] Step S1013, performing prior knowledge annotation of the sample significant feature data on part of the sample IoT session data obtained in the second IoT service environment to generate the second sample IoT session data sequence.

[0058] The server uses a similar processing method for the data of the second IoT service environment. However, considering resource limitations or differences in data processing strategies, the server may only perform detailed prior knowledge annotation on part of the data. These selected data undergo the same analysis process as the data of the first IoT service environment, including identifying significant features and adding prior knowledge labels. Finally, the server generates a second sample IoT session data sequence, which provides important reference information for subsequent neural network training and data desensitization.

[0059] In a possible implementation, step S103 may include: Step S1031: Generate a first significant feature error based on the feature distance between the sample significant feature data in the first desensitized data sequence and the sample significant feature data in the first sample IoT session data sequence.

[0060] In this embodiment, the server first obtains the first desensitized data sequence and the first example IoT session data sequence. These first desensitized data sequence and the first example IoT session data sequence both contain sample significant feature data, that is, key information points in the data, such as specific communication modes, data transmission volume peaks, etc.

[0061] In order to evaluate the impact of desensitization on the significant features in the data, the server uses a feature distance calculation method (such as Euclidean distance, cosine similarity, etc.) to quantify the difference between the sample significant feature data in the first desensitized data sequence and the corresponding feature data in the first sample IoT session data sequence. The difference value, namely the feature distance, reflects the degree of retention of the significant features after desensitization.

[0062] After the server calculates this feature distance, it defines it as the first significant feature error. The smaller the first significant feature error is, the smaller the impact of the desensitization processing on the significant features in the first example IoT session data is, and the better the desensitization effect is.

[0063] Step S1032: Generate a second significant feature error based on the feature distance between the sample significant feature data in the second desensitized data sequence and the sample significant feature data in the second sample IoT session data sequence.

[0064] Similar to step S1031, the server then obtains the second desensitized data sequence and the second sample IoT session data sequence. Similarly, these second desensitized data sequence and the second sample IoT session data sequence also contain sample significant feature data.

[0065] The server uses the same feature distance calculation method to quantify the difference between the sample significant feature data in the second desensitized data sequence and the corresponding feature data in the second sample IoT session data sequence. The calculated feature distance is defined as the second significant feature error.

[0066] Step S1033: generating the significant feature error based on the first significant feature error and the second significant feature error.

[0067] After obtaining the first significant feature error and the second significant feature error, the server needs to conduct a comprehensive analysis of the two errors to evaluate the effect of the entire desensitization process.

[0068] The server can use a variety of methods to generate the final significant feature error, such as taking the average, weighted average, maximum value, etc. of the two errors. The comprehensive error value more comprehensively reflects the impact of the desensitization processing on the significant features of the data in two different IoT service environments.

[0069] For example, the server can calculate the weighted sum of the first significant feature error and the second significant feature error as the final significant feature error, and the weighted sum can be adjusted according to factors such as the importance and data volume of the two IoT service environments. In this way, the server obtains a significant feature error value that can fully reflect the effect of the desensitization process.

[0070] In a possible implementation, before step S104, the method further includes: Step A110: Load the first desensitized data sequence into the second desensitization prediction unit to perform data desensitization prediction to generate a first reconstructed session data sequence for the first Internet of Things service environment.

[0071] In this embodiment, the server first loads the first desensitized data sequence, which has been processed by the first desensitization prediction unit and has removed sensitive information, into the second desensitization prediction unit. The task of the second desensitization prediction unit is to try to reconstruct the desensitized data to predict its original form. This process is similar to a decoding process, which aims to verify whether the desensitized data retains enough information for reconstructing the original data.

[0072] After being processed by the second desensitizing prediction unit, the server obtains a set of reconstructed session data sequences, which are called the first reconstructed session data sequences of the first IoT service environment. These first reconstructed session data sequences are original data forms predicted based on the desensitized data, and are used for subsequent comparison with the original sample data to evaluate the effects of desensitization and reconstruction.

[0073] Step A120: Load the second desensitized data sequence into the first desensitization prediction unit to perform data desensitization prediction to generate a second reconstructed session data sequence for the second Internet of Things service environment.

[0074] Similar to step A110, the server then loads the second desensitized data sequence into the first desensitization prediction unit for data desensitization prediction. These second desensitized data sequences come from the second Internet of Things service environment and have been desensitized by the second desensitization prediction unit.

[0075] Through processing by the first desensitized prediction unit, the server obtains a second reconstructed session data sequence of the second Internet of Things service environment. These second reconstructed session data sequences are original data forms predicted based on the second desensitized data sequences, and are also used for subsequent comparative evaluation with the original sample data.

[0076] Step A130: Generate a first reconstruction error parameter based on the correlation between the first reconstructed session data sequence and the first sample IoT session data sequence.

[0077] The server compares the first reconstructed session data sequence with the original first sample IoT session data sequence, and evaluates the degree of consistency between the reconstructed data and the original data by calculating the correlation between the two (such as using indicators such as correlation coefficient and mean square error). The correlation indicator is defined as the first reconstruction error parameter, which reflects the degree of retention of the original data during the desensitization and reconstruction process.

[0078] Step A140: Generate a second reconstruction error parameter based on the correlation between the second reconstructed session data sequence and the second sample IoT session data sequence.

[0079] Similarly, the server also compares the second reconstructed session data sequence with the original second sample IoT session data sequence, and calculates a correlation index between the two, which is defined as a second reconstruction error parameter.

[0080] Step A150: generating the reconstruction error parameter based on the first reconstruction error parameter and the second reconstruction error parameter.

[0081] The server comprehensively processes the first reconstruction error parameter and the second reconstruction error parameter, such as taking an average, a weighted sum, or other statistical methods, to obtain an overall reconstruction error parameter, which more comprehensively reflects the effects of the desensitization and reconstruction processes in two different IoT service environments.

[0082] Step S104 includes: generating a global error parameter corresponding to the candidate neural network based on the reconstruction error parameter and the significant feature error, and training the candidate neural network based on the global error parameter.

[0083] After obtaining the reconstruction error parameter and the previously calculated significant feature error, the server combines these two error parameters to generate a global error parameter. The global error parameter comprehensively considers the degree of retention of significant features in the desensitization process and the accuracy of the reconstructed data, and is an important indicator for evaluating the performance of candidate neural networks.

[0084] Finally, the server uses this global error parameter to train the candidate neural network. Through optimization techniques such as the back-propagation algorithm, the weights and parameters of the network are adjusted to minimize the global error parameter. This process aims to improve the ability of the neural network to retain significant features and reconstruct the original data during the desensitization process.

[0085] In a possible implementation, the candidate neural network further includes a first identification unit for identifying the validity of the session log on the second Internet of Things service environment, and a second identification unit for identifying the validity of the session log on the first Internet of Things service environment.

[0086] Before step S104, the method further includes: Step B110, loading the first desensitized data sequence into the first identification unit to generate a first identification result, loading the second desensitized data sequence into the second identification unit to generate a second identification result, loading the second sample IoT session data sequence into the first identification unit to generate a third identification result, loading the first sample IoT session data sequence into the second identification unit to generate a fourth identification result.

[0087] In this embodiment, in addition to the units for data desensitization and reconstruction, the candidate neural network is also specially designed with two identification units. The first identification unit is designed to identify the validity of the session log in the second IoT service environment, while the second identification unit is used to identify the validity of the session log in the first IoT service environment. These two identification units can identify abnormal or invalid logs by analyzing the characteristics and data patterns of the session logs, thereby improving the reliability and security of the data.

[0088] Before training the candidate neural network, the server needs to load different data sequences into the corresponding identification unit for processing.

[0089] First, the server loads the first desensitized data sequence into the first identification unit. These data are desensitized and are intended to test the identification ability of the first identification unit for the desensitized data. After processing, the first identification unit outputs a first identification result, which reflects the validity of the session log in the first desensitized data sequence.

[0090] Next, the server loads the second desensitized data sequence into the second identification unit and obtains a second identification result. This process is to verify the accuracy of the second identification unit in identifying the desensitized data from the second Internet of Things service environment.

[0091] In order to evaluate the authentication capability of the authentication unit for the original sample data, the server also loads the second sample IoT session data sequence into the first authentication unit and generates a third authentication result. Similarly, the first sample IoT session data sequence is loaded into the second authentication unit and generates a fourth authentication result.

[0092] Step B120, generating a first identification comparison error parameter based on the first identification result and the third identification result, generating a second identification comparison error parameter based on the second identification result and the fourth identification result, and generating an identification comparison error parameter based on the first identification comparison error parameter and the second identification comparison error parameter.

[0093] After obtaining four authentication results, the server begins error analysis.

[0094] First, the server compares the first identification result and the third identification result. Since the first identification result is based on the desensitized data, and the third identification result is based on the original sample data, by comparing the two, the impact of the desensitization process on the identification result can be evaluated. The server calculates the difference between the two and defines it as the first identification comparison error parameter.

[0095] Next, the server compares the second authentication result and the fourth authentication result, and similarly calculates the difference between the two, and defines the difference as a second authentication comparison error parameter.

[0096] Finally, the server combines the first identification comparison error parameter and the second identification comparison error parameter, and generates an overall identification comparison error parameter through certain statistical methods (such as weighted average, standard deviation analysis, etc.). The identification comparison error parameter reflects the performance difference of the identification unit when processing desensitized data and original data, and provides an important reference for the subsequent neural network training.

[0097] In a possible implementation, generating a global error parameter corresponding to the candidate neural network based on the reconstruction error parameter and the significant feature error includes: Step S1041, obtaining the training participation coefficient of the reconstruction error parameter and the training participation coefficient of the significant feature error.

[0098] Step S1042: Based on the training participation coefficient of the reconstruction error parameter and the training participation coefficient of the significant feature error, the reconstruction error parameter and the significant feature error are fused to generate a fused error parameter.

[0099] Step S1043, generating a global error parameter corresponding to the candidate neural network based on the fusion error parameter and the identification and comparison error parameter.

[0100] In this embodiment, before training the candidate neural network, the server needs to first determine the weights of different error parameters in the training process, which involves two important training participation coefficients: the training participation coefficient of the reconstruction error parameter and the training participation coefficient of the significant feature error.

[0101] These two coefficients are usually determined by experts based on experience or through experiments, and they reflect the relative contribution of reconstruction error and significant feature error to the global error during training. For example, if the reconstruction error is considered more important, its training participation coefficient can be set relatively high.

[0102] The server obtains these training participation coefficients from the configuration file, database, or user input to prepare for subsequent error fusion.

[0103] After obtaining the reconstruction error parameters and the training participation coefficients of the significant feature errors, the server starts error fusion.

[0104] Specifically, the server first multiplies the reconstruction error parameter and the salient feature error by their respective training participation coefficients, so that the errors can be weighted according to their importance. Then, the server adds these two weighted errors to generate a fusion error parameter.

[0105] This fusion error parameter comprehensively considers the reconstruction quality and the retention of significant features, and is an important indicator for evaluating the performance of candidate neural networks.

[0106] The discrimination comparison error parameter reflects the performance difference of the discrimination unit when processing the desensitized data and the original data. In order to take this error parameter into consideration of the global error, the server can combine it with the fusion error parameter.

[0107] Specifically, the server can perform weighted summation of the discrimination comparison error parameter and the fusion error parameter to generate a global error parameter, which comprehensively considers the reconstruction quality, the retention of significant features, and the performance of the discrimination unit, and is an important basis for evaluating and optimizing candidate neural networks.

[0108] After obtaining the global error parameter, the server can use this parameter to train and optimize the candidate neural network. By continuously adjusting the weights and parameters of the network and training with the goal of minimizing the global error parameter, the comprehensive capabilities of the neural network in desensitization, data reconstruction, and identification of log validity can be improved.

[0109] In a possible implementation, the training of the candidate neural network based on the global error parameter includes: Step S1044: lock the weight configuration information of the first identification unit and the second identification unit, and train the first desensitization prediction unit and the second desensitization prediction unit based on the global error parameter.

[0110] Step S1045 , locking the weight configuration information of the first desensitization prediction unit and the second desensitization prediction unit, and training the first identification unit and the second identification unit based on the identification comparison error parameter.

[0111] In this embodiment, during the training process, the server first locks the weight configuration information of the first identification unit and the second identification unit, which means that during this part of the training process, the weights of the two identification units will not be updated or changed. The purpose of locking the weights is to optimize the desensitization prediction unit and the identification unit separately, rather than optimizing all units at the same time, so that the training process can be more focused and targeted.

[0112] After locking the weights of the discriminator, the server will focus on training the first desensitization prediction unit and the second desensitization prediction unit. At this time, the server can use the previously calculated global error parameter as a training guide. The global error parameter comprehensively considers the reconstruction quality, the retention of significant features, and the performance of the discriminator, so it is a good training indicator.

[0113] The server passes the global error parameter back to the desensitization prediction unit through the back propagation algorithm, and adjusts the weights of the first desensitization prediction unit and the second desensitization prediction unit according to this error. Through multiple iterative training, the server aims to reduce the global error and thus improve the performance of the desensitization prediction unit.

[0114] After completing the training of the desensitization prediction unit, the server can lock the weights of the two units. The purpose of doing this is to focus on optimizing the performance of the identification unit in the subsequent training without affecting the previously optimized desensitization prediction unit.

[0115] After locking the weight of the desensitization prediction unit, the server will use the discrimination comparison error parameter to train the first discrimination unit and the second discrimination unit. The discrimination comparison error parameter reflects the performance difference of the discrimination unit when processing desensitized data and original data, and is therefore a key indicator for optimizing the discrimination unit.

[0116] The server also transmits the identification comparison error parameter back to the identification unit through the back propagation algorithm, and adjusts the weights of the first identification unit and the second identification unit according to the error. Through multiple iterative training, the server aims to reduce the identification comparison error, thereby improving the identification unit's ability to identify the validity of the session log.

[0117] Through this phased training method, the server can optimize the performance of the desensitization prediction unit and the identification unit respectively, and finally obtain a candidate neural network that can not only effectively desensitize and reconstruct data, but also accurately identify the validity of session logs.

[0118] In a possible implementation, step S130 may include: obtaining a target session log of the second Internet of Things service environment generated by data desensitization by a first desensitization prediction unit of the Internet of Things session desensitization network.

[0119] In a possible implementation, the significant feature data included in the target session log is a feature part marked for the privacy access control task. The step of optimizing the privacy access control network of the second Internet of Things service environment based on the target session log of the second Internet of Things service environment includes: Step S141, obtaining a privacy access control network of a second Internet of Things service environment, wherein the privacy access control network is used to perform the privacy access control task according to the significant feature data.

[0120] Step S142: Acquire a sample learning data sequence, where the sample learning data sequence includes a target session log and annotation data for indicating the significant feature data.

[0121] Step S143: Optimize the privacy access control network based on the sample learning data sequence.

[0122] In this embodiment, in the scenario of the smart home system, the server communicates with the smart center in the home to obtain the deployed privacy access control network, which is specially designed for the smart home system to ensure the privacy and data security of family members. It is responsible for controlling the information access rights between smart devices to prevent unauthorized access and data leakage.

[0123] In the scenario of the industrial Internet of Things environment, the server is connected to the factory's control center to obtain the privacy access control network that has been established inside the factory. This network is used to protect the security of industrial equipment and production data, ensuring that only authorized personnel can access sensitive information and control systems.

[0124] Furthermore, in the smart home system, the server extracts sample learning data sequences from the desensitized target session logs. These data sequences record the interaction information between family members and smart devices, such as lighting control, temperature adjustment, security system deployment, etc. At the same time, these data also contain annotation information, indicating which data is privacy-sensitive information and needs special protection, such as family member identity information, device control passwords, etc.

[0125] In the industrial IoT environment, the server obtains sample learning data sequences from the desensitized target session logs, which record the operating status of industrial equipment, production data, and related control instructions. The labeled data indicates which information is critical and needs to be protected, such as production formulas, equipment maintenance records, etc.

[0126] Furthermore, in the smart home system, the server uses the extracted sample learning data sequence to optimize the privacy access control network. By comparing the network output results with the labeled data, the server adjusts the network parameters and rules to improve the accuracy and efficiency of privacy access control. In this way, when family members interact with smart devices, the privacy access control network can more accurately determine whether the access request is legitimate and take corresponding protection measures.

[0127] In the industrial IoT environment, the server uses sample learning data sequences to train and optimize the privacy access control network. By simulating various access requests and attack scenarios, the server continuously adjusts the network's defense strategies and access rules to ensure that only legitimate users can access sensitive data and control systems, which helps improve the factory's information security level and prevent malicious attacks and data leaks.

[0128] Figure 2 The privacy access control system 100 in the IoT environment shown includes: a processor 1001 and a memory 1003. The processor 1001 and the memory 1003 are connected, such as through a bus 1002. Optionally, the privacy access control system 100 in the IoT environment may also include a transceiver 1004, which may be used for data interaction between the server and other servers, such as data transmission and / or data reception. It should be noted that the transceiver 1004 is not limited to one in actual scheduling, and the structure of the privacy access control system 100 in the IoT environment does not constitute a limitation on the embodiments of the present application.

[0129] Processor 1001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It may implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 1001 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and the like.

[0130] The bus 1002 may include a path to transmit information between the above components. The bus 1002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 1002 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 2 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0131] The memory 1003 may be a ROX (Read Only Xemory) or other types of static storage devices that can store static information and instructions, a RAX (Random Access Xemory) or other types of dynamic storage devices that can store information and instructions, or an EEPROX (Electrically Erasable Programmable Read Only Xemory), a CD-ROX (Compact Disc Read Only Xemory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), magnetic disk storage media, other magnetic storage devices, or any other medium that can be used to carry or store program code and can be read by a computer, without limitation herein.

[0132] The memory 1003 is used to store program codes for executing the embodiments of the present application, and the execution is controlled by the processor 1001. The processor 1001 is used to execute the program codes stored in the memory 1003 to implement the steps shown in the above method embodiments.

[0133] An embodiment of the present application provides a computer-readable storage medium having program code stored thereon. When the program code is executed by a processor, the steps and corresponding contents of the aforementioned method embodiment can be implemented.

[0134] It should be understood that, although each operation step is indicated by arrows in the flowchart of the embodiment of the present application, the implementation order of these steps is not limited to the order indicated by the arrows. Unless there is clear explanation in this article, in some implementation scenarios of the embodiment of the present application, the implementation steps in each flowchart can be performed in other orders based on demand. In addition, some or all of the steps in each flowchart may include multiple sub-steps or multiple stages according to the actual implementation scenario, and some or all of these sub-steps or stages may be executed at the same time, and each sub-step or stage in these sub-steps or stages may also be executed at different times respectively. In different scenarios at the execution time, the execution order of these sub-steps or stages can be flexibly configured based on demand, and the embodiment of the present application does not limit this.

[0135] The above is only an optional implementation method for some implementation scenarios of the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the technical concept of the scheme of the present application, other similar implementation methods based on the technical ideas of the present application are also within the protection scope of the embodiments of the present application.

Claims

1. A privacy access control method in an Internet of Things environment, characterized in that: The method comprises: Obtain the IoT session log of the first IoT service environment; Loading the IoT session log into an IoT session desensitization network, wherein the IoT session desensitization network is generated by network learning based on significant feature errors, wherein the significant feature errors are determined based on sample significant feature data in a first desensitized data sequence generated after data desensitization of a first sample IoT session data sequence in a first IoT service environment, and sample significant feature data in a second desensitized data sequence generated after data desensitization of a second sample IoT session data sequence in a second IoT service environment; Obtaining a target session log of the second Internet of Things service environment generated by the Internet of Things session desensitization network performing data desensitization on the Internet of Things session log, wherein the significant feature data included in the target session log matches the significant feature data included in the Internet of Things session log; After optimizing the privacy access control network of the second Internet of Things service environment based on the target session log of the second Internet of Things service environment, the target privacy access control task is performed based on the optimized privacy access control network.

2. The privacy access control method in the Internet of Things environment according to claim 1 is characterized in that: Before loading the IoT session log into the IoT session desensitization network, the method further includes: Acquire the first sample IoT session data sequence, the second sample IoT session data sequence, and a candidate neural network, wherein the candidate neural network includes a first desensitization prediction unit and a second desensitization prediction unit, and the sample IoT session data in the first sample IoT session data sequence and the second sample IoT session data sequence are both annotated with prior label knowledge for indicating the sample significant feature data; The first sample IoT session data sequence is loaded into the first desensitization prediction unit, and data desensitization is performed to generate a first desensitized data sequence of the second IoT service environment annotated with the sample significant feature data, and the second sample IoT session data sequence is loaded into the second desensitization prediction unit, and data desensitization is performed to generate a second desensitized data sequence of the first IoT service environment annotated with the sample significant feature data; Determining the significant feature error based on the sample significant feature data in the second desensitized data sequence and the sample significant feature data in the first desensitized data sequence; The candidate neural network is trained based on the significant feature error to generate the Internet of Things session desensitization network.

3. The privacy access control method in the Internet of Things environment according to claim 2 is characterized in that: The obtaining of the first sample IoT session data sequence, the second sample IoT session data sequence, and the candidate neural network includes: Obtaining sample IoT session data of the first IoT service environment, and obtaining sample IoT session data of the second IoT service environment; Performing prior knowledge annotation of the sample significant feature data on the acquired sample IoT session data of the first IoT service environment to generate the first sample IoT session data sequence; Prior knowledge annotation of the sample significant feature data is performed on part of the sample Internet of Things session data obtained in the second Internet of Things service environment to generate the second sample Internet of Things session data sequence.

4. The privacy access control method in the Internet of Things environment according to claim 2 is characterized in that: The determining the significant feature error based on the sample significant feature data in the second desensitized data sequence and the sample significant feature data in the first desensitized data sequence includes: Generate a first significant feature error based on a feature distance between sample significant feature data in the first desensitized data sequence and sample significant feature data in the first sample IoT session data sequence; Generate a second significant feature error based on a feature distance between sample significant feature data in the second desensitized data sequence and sample significant feature data in the second sample IoT session data sequence; The significant feature error is generated based on the first significant feature error and the second significant feature error.

5. The privacy access control method in the Internet of Things environment according to claim 2 is characterized in that: Before training the candidate neural network based on the significant feature error, the method further includes: Loading the first desensitized data sequence into the second desensitization prediction unit to perform data desensitization prediction to generate a first reconstructed session data sequence for the first Internet of Things service environment; Loading the second desensitized data sequence into the first desensitization prediction unit to perform data desensitization prediction, and generating a second reconstructed session data sequence for the second Internet of Things service environment; Generate a first reconstruction error parameter based on the correlation between the first reconstructed session data sequence and the first sample IoT session data sequence; generating a second reconstruction error parameter based on the correlation between the second reconstructed session data sequence and the second sample IoT session data sequence; generating the reconstruction error parameter based on the first reconstruction error parameter and the second reconstruction error parameter; The training of the candidate neural network based on the significant feature error comprises: A global error parameter corresponding to the candidate neural network is generated based on the reconstruction error parameter and the significant feature error, and the candidate neural network is trained based on the global error parameter.

6. The privacy access control method in the Internet of Things environment according to claim 5 is characterized in that: The candidate neural network further includes a first authentication unit for authenticating the validity of the session log on the second IoT service environment, and a second authentication unit for authenticating the validity of the session log on the first IoT service environment; Before training the candidate neural network based on the global error parameter, the method further includes: Loading the first desensitized data sequence into the first identification unit to generate a first identification result; Loading the second desensitized data sequence into the second identification unit to generate a second identification result; Loading the second sample IoT session data sequence into the first authentication unit to generate a third authentication result; Loading the first sample IoT session data sequence into the second authentication unit to generate a fourth authentication result; generating a first authentication comparison error parameter based on the first authentication result and the third authentication result; generating a second authentication comparison error parameter based on the second authentication result and the fourth authentication result; generating an identification comparison error parameter based on the first identification comparison error parameter and the second identification comparison error parameter; The generating a global error parameter corresponding to the candidate neural network based on the reconstruction error parameter and the significant feature error comprises: Obtaining a training participation coefficient of the reconstruction error parameter and a training participation coefficient of the significant feature error; Based on the training participation coefficient of the reconstruction error parameter and the training participation coefficient of the significant feature error, fusing the reconstruction error parameter and the significant feature error to generate a fusion error parameter; A global error parameter corresponding to the candidate neural network is generated based on the fusion error parameter and the discrimination comparison error parameter.

7. The privacy access control method in the Internet of Things environment according to claim 6 is characterized in that: The training of the candidate neural network based on the global error parameter comprises: Locking the weight configuration information of the first identification unit and the second identification unit, and training the first desensitization prediction unit and the second desensitization prediction unit based on the global error parameter; The weight configuration information of the first desensitization prediction unit and the second desensitization prediction unit is locked, and the first identification unit and the second identification unit are trained based on the identification comparison error parameter.

8. The privacy access control method in the Internet of Things environment according to claim 7 is characterized in that: The step of obtaining the target session log of the second Internet of Things service environment generated by performing data desensitization on the Internet of Things session log by the Internet of Things session desensitization network includes: Obtain a target session log of the second Internet of Things service environment generated by data desensitization performed by the first desensitization prediction unit of the Internet of Things session desensitization network.

9. The privacy access control method in an Internet of Things environment according to any one of claims 1 to 8, characterized in that: The significant feature data included in the target session log is a feature part marked for the privacy access control task; the step of optimizing the privacy access control network of the second Internet of Things service environment based on the target session log of the second Internet of Things service environment includes: Acquire a privacy access control network of a second Internet of Things service environment, wherein the privacy access control network is used to perform the privacy access control task according to the significant feature data; Acquire a sample learning data sequence, wherein the sample learning data sequence includes a target session log and annotation data indicating the significant feature data; The privacy access control network is optimized based on the sample learning data sequence.

10. A privacy access control system in an Internet of Things environment, characterized in that: It includes a processor and a computer-readable storage medium, wherein the computer-readable storage medium stores machine-executable instructions, and when the machine-executable instructions are executed by the processor, the privacy access control method in the Internet of Things environment described in any one of claims 1 to 9 is implemented.

Citation Information

Patent Citations

  • Deep learning-oriented data sensitive attribute desensitization system and method

    CN114417427A

  • Access processing method and system fusing structured and unstructured data

    CN117033433A

  • Network intrusion detection method and device, electronic device and storage medium

    CN117354046A

  • Data processing method and system applied to network service

    CN117785964A

  • Enterprise big data mining method and system based on artificial intelligence

    CN118094639A

Cited By

  • Secure communication method, system and device based on big data and medium

    CN119675981A