Internet of Things service data encryption and privacy protection method and device

Abstract

The invention relates to the technical field of government affair Internet of Things, and discloses an Internet of Things service data encryption and privacy protection method and device, and the method comprises the steps: collecting real-time data items from government affair Internet of Things equipment through a wireless sensor network through the Internet of Things technology to form a government affair data set, and carrying out the encryption processing; calculating and evaluating the sensitivity delta f of each data item in the government affair data set and a query result corresponding to the data item through a differential privacy algorithm, and adding random noise according to the sensitivity delta f and a privacy budget epsilon, and performing government affair service efficiency, government affair anomaly monitoring, government affair demand prediction and government affair complaint risk prediction analysis on the government affair data set added with the random noise by adopting a machine learning algorithm, providing data service for an analysis result through an API (Application Program Interface), and dynamically adjusting the sensitivity delta f, the privacy budget epsilon and the adding degree of the random noise according to an evaluation result. While data security and government affair privacy are guaranteed, data analysis efficiency and accuracy are improved, government affair management is optimized, and adaptability and expandability are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of government Internet of Things, and in particular to a method and device for encrypting and protecting privacy of Internet of Things business data. Background Art

[0002] With the rapid development of smart government and Internet of Things technology, government Internet of Things business has become an indispensable part of modern government systems. Through Internet of Things technologies such as wireless sensor networks, government Internet of Things has realized real-time monitoring and data collection of government equipment, providing strong support for the intelligent management of government affairs. However, with the continuous expansion of government Internet of Things business, the data security and privacy protection issues it faces are becoming increasingly prominent. Data encryption is the core means to protect the data security of Internet of Things business. By encrypting sensitive data, unauthorized access and theft can be effectively prevented, ensuring the integrity and confidentiality of data during transmission and storage. In the government Internet of Things, a large amount of real-time data needs to be collected, processed and released to relevant stakeholders, such as government service providers, information management companies, government regulatory superiors, etc. Government Internet of Things business involves a large amount of user data and sensitive information, such as government applicant information, government applicant application matters, government information volume, government information type, equipment status, etc. Once this information is leaked, it will pose a serious threat to government privacy. Therefore, privacy protection is particularly important in government Internet of Things business. Summary of the invention

[0003] In order to solve the above problems, the present invention provides a method and device for encrypting and protecting privacy of Internet of Things business data, which not only ensures data security and government privacy, but also improves the efficiency and accuracy of data analysis, optimizes government management, enhances the security and flexibility of data services, and has good adaptability and scalability.

[0004] The present invention provides an Internet of Things service data encryption and privacy protection method, which specifically includes the following steps:

[0005] S1. Through the Internet of Things technology, wireless sensor networks are used to collect real-time data items from government Internet of Things devices and transmit them to the cloud computing platform to form a government data set, and the government data set is encrypted;

[0006] S2. Calculate the maximum possible change of each data item and the query result corresponding to the data item in the government data set through the differential privacy algorithm, and evaluate the sensitivity Δf of each data item and the query result corresponding to the data item in the government data set;

[0007] S3, adding random noise to the data item and the query result corresponding to the data item according to the sensitivity Δf and the privacy budget ε of each data item and the query result corresponding to the data item in the government data set;

[0008] S4. Use machine learning algorithms to perform government service efficiency analysis, government anomaly monitoring analysis, government demand forecasting analysis, and government complaint risk forecasting analysis on government data sets with random noise added;

[0009] S5. Provide data services through API interfaces with analysis results, adopt strong identity authentication mechanisms, set access control rules based on user roles and permissions, and set access rights to each data item in the government data set and the query results corresponding to the data item through the access control list ACL;

[0010] S6. Evaluate the analysis results and query results in S4 and S5, and dynamically adjust the sensitivity Δf, privacy budget ε, and the degree of random noise addition according to the evaluation results.

[0011] Furthermore, the S1 specifically includes: the government Internet of Things devices include self-service devices, including intelligent guide robots and self-service terminals of government service agencies, which collect real-time data items including but not limited to government applicant information, government information type, service processing volume, processing time, processing results, public satisfaction evaluation, complaint and suggestion data, etc., and encrypt the government data set of the government Internet of Things through encryption algorithms and key generation, storage, distribution and update mechanisms.

[0012] Furthermore, the sensitivity Δf of evaluating each data item in the government data set and the query result corresponding to the data item in S2 is specifically as follows: for non-numeric data, the square root of the sum of squares of changes in the query result corresponding to the data item caused by the change in the data item is calculated to obtain the sensitivity Δf of each data item in the government data set and the query result corresponding to the data item, and scores are assigned to different query results, and they are sorted according to the scores. The query result with a higher score has a higher sensitivity Δf;

[0013] For numerical data, the maximum absolute change in the query result corresponding to the data item when the data item changes is calculated, Δf = max_{x,x'∈D}|f(x)-f(x')|, where D is the data set, x and x' are two adjacent data items in the data set, f is the query function, and Δf is the maximum change in the query result;

[0014] The higher the sensitivity Δf of a data item or the query result corresponding to the data item, the more random noise needs to be added.

[0015] Furthermore, the random noise is added to the data items and the query results corresponding to the data items in S3 specifically as follows: for numerical data, noise that obeys the Laplace distribution Lap(Δf / ε) is added to the query results of the data items and the data items; for non-numerical data, the scores of different query results are evaluated and sorted by a scoring function, and the probability distribution of the output query results is adjusted according to the sorting results and the privacy budget ε.

[0016] Furthermore, the adding random noise to the numerical data in S3 further includes:

[0017] S31. According to the sensitivity Δf and the privacy budget ε calculated by the differential privacy algorithm, the scale parameter of the noise is determined to be Δf / ε.

[0018] S32, using a random number generator to generate random noise that obeys Laplace distribution, and randomly extracting a noise value;

[0019] S33. Add the generated noise value to the government data set or query result.

[0020] Furthermore, the S4 specifically includes:

[0021] Government service efficiency analysis: By analyzing historical data, we can identify the changing trends of government service demand and response efficiency, and provide a basis for future government planning and scheduling;

[0022] Government affairs abnormality monitoring: Real-time monitoring of abnormal events in government affairs, including security event monitoring, business processing abnormalities, etc., timely discovery and resolution of problems to ensure the safe and stable operation of government affairs;

[0023] Government demand forecasting: Based on historical data and current trends, forecast future government demand and provide guidance for government item addition and allocation;

[0024] Government complaint risk prediction: Build a government complaint risk scoring model through government data and communication data to improve the accuracy and efficiency of arrears prediction.

[0025] Furthermore, the S4 specifically includes: dividing the government data set into a government training set, a government verification set and a government test set, the government training set is used to train the model, the government verification set is used to adjust the model parameters, and the government test set is used to evaluate the model performance, using the government training set to train the model, adjusting the model parameters by minimizing the prediction error, and during the training process, using cross-validation and other techniques to prevent overfitting, using the government test set to evaluate the trained model, calculating indicators such as prediction accuracy and recall rate to measure the performance of the model, adjusting the model parameters or selecting other models for trial based on the evaluation results, wherein a linear regression model is used to predict continuous numerical data, such as government response prediction, a decision tree is used to perform classification and regression tasks, such as government fault diagnosis, and a neural network model is used to train nonlinear problems, such as government consumption optimization.

[0026] Furthermore, the S6 specifically includes:

[0027] Sensitivity adjustment: Re-evaluate the sensitivity of data items and query results based on changes in data and business requirements, and adjust their values ​​accordingly.

[0028] Privacy budget adjustment: According to the privacy protection requirements and the accuracy requirements of data analysis, adjust the value of the privacy budget ε. The larger the privacy budget, the less noise is added, but the lower the privacy protection; the smaller the privacy budget, the more noise is added, but the higher the privacy protection.

[0029] Adjust the degree of noise addition: Recalculate and add random noise based on the adjusted sensitivity and privacy budget to ensure data privacy and analysis accuracy.

[0030] Furthermore, the S5 specifically includes: the strong authentication mechanism is a multi-factor authentication mechanism, including username / password, dynamic password, biometrics, etc., and the access control rules set according to user roles and permissions are specifically divided into users into different roles, and a set of permissions are defined for each role, and access rules are set according to role permissions: Rule 1: Administrators can access all data items and perform all operations, including reading, writing, modifying, and deleting; Rule 2: Data analysts can read specific data items, including power statistics data, query results in response to prediction results, but cannot modify or delete data; Rule 3: Operation and maintenance personnel can access system logs and monitoring data, but cannot access business data; Rule 4: Ordinary users are authorized to access limited data items or query results based on their business needs.

[0031] The present invention also provides an Internet of Things service data encryption and privacy protection device, which uses the Internet of Things service data encryption and privacy protection method as described in any one of claims 1 to 9, and is characterized in that it includes:

[0032] The data collection and encryption processing unit uses the Internet of Things technology to collect real-time data items from government Internet of Things devices using a wireless sensor network and transmits them to the cloud computing platform to form a government data set. The government data set is encrypted and then output to the data sensitivity assessment unit;

[0033] The data sensitivity evaluation unit calculates the maximum possible change of each data item in the government data set and the query result corresponding to the data item through a differential privacy algorithm, evaluates the sensitivity Δf of each data item in the government data set and the query result corresponding to the data item, and the sensitivity Δf evaluation result is transferred to the data noise adding unit;

[0034] A data noise adding unit, which adds random noise to the data item and the query result corresponding to the data item according to the sensitivity Δf and the privacy budget ε of each data item in the government data set and the query result corresponding to the data item, and outputs the government data set after adding the random noise to the government data set analysis unit for analysis;

[0035] The government data set analysis unit uses a machine learning algorithm to perform government service efficiency analysis, government anomaly monitoring analysis, government demand forecasting analysis, and government complaint risk forecasting analysis on the government data set after random noise is added, and the analysis results are output to the interaction and authority control unit and the dynamic adjustment unit;

[0036] The interaction and permission control unit provides data services through the API interface with the analysis results, adopts a strong identity authentication mechanism, and sets access control rules according to user roles and permissions. It sets access rights to each data item in the government data set and the query results corresponding to the data item through the access control list ACL, and the query results are fed back to the dynamic adjustment unit;

[0037] The dynamic adjustment unit evaluates the analysis results in the government data set analysis unit and the query results in the interaction and permission control unit, dynamically adjusts the sensitivity Δf, privacy budget ε and the degree of random noise addition according to the evaluation results, and feeds back to the data sensitivity evaluation unit and the data noise addition unit.

[0038] The beneficial effects of the present invention are:

[0039] Enhanced data security: Government data is collected through IoT technology and encrypted on the cloud computing platform, effectively preventing the risk of data leakage during transmission and storage. The use of encryption algorithms and key management mechanisms ensures the security of government data sets and prevents unauthorized access and tampering.

[0040] Protecting government privacy: A differential privacy algorithm is introduced to evaluate the sensitivity of each data item and query result in the government data set, and random noise is added accordingly, thereby allowing effective analysis of the data while protecting government privacy. By dynamically adjusting the sensitivity, privacy budget, and the degree of noise addition, the needs of privacy protection and data analysis are balanced, ensuring the compliance of data use and the privacy rights of users.

[0041] Improve data analysis efficiency and accuracy: Use machine learning algorithms to perform multi-dimensional analysis on encrypted and noisy government data sets, including government service efficiency analysis, anomaly monitoring, demand forecasting, and government complaint risk forecasting, which improves the efficiency and accuracy of data analysis. By dividing the government training set, validation set, and test set, and using multiple models for training and optimization, the accuracy and robustness of data analysis are further improved.

[0042] Optimize government administration: Based on the analysis of government service efficiency and demand forecast results, it provides a scientific basis for government planning and scheduling, helps optimize government resource allocation, and improves government operation efficiency. Real-time monitoring of abnormal events in government affairs and rapid response and processing ensures the safe and stable operation of government affairs.

[0043] Enhance the security and flexibility of data services: Data services are provided through API interfaces, and strong authentication mechanisms and multi-factor authentication are adopted to ensure the security of data services. Access control rules are set according to user roles and permissions, and fine-grained access control is implemented through ACL, which not only ensures the compliance of data use, but also meets the business needs of different users.

[0044] Adaptability and scalability: This method can dynamically adjust the sensitivity and privacy budget according to changes in data and business needs, ensuring the flexibility and adaptability of data protection and analysis strategies. At the same time, this method has good scalability and can be further optimized and improved with the continuous development of government IoT technology to adapt to more complex data protection and analysis needs.

[0045] In summary, the IoT business data encryption and privacy protection method of the present invention improves the efficiency and accuracy of data analysis, optimizes government management, enhances the security and flexibility of data services, and has good adaptability and scalability while ensuring data security and government privacy. By integrating multiple functional units, comprehensive protection, security analysis and dynamic adjustment of government IoT data are achieved, providing strong support for the safe operation and efficient management of smart government affairs. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a flow chart of a method for encrypting and protecting privacy of IoT business data;

[0047] Figure 2 The figure is a schematic diagram of an IoT business data encryption and privacy protection device. DETAILED DESCRIPTION

[0048] The following is combined with Figure 1-2 The preferred embodiments of the present invention are described in detail so that the advantages and features of the present invention can be more easily understood by those skilled in the art, thereby making a clearer and more definite definition of the protection scope of the present invention.

[0049] First, through the Internet of Things technology, wireless sensor networks are used to collect real-time data items from government Internet of Things devices and transmit them to the cloud computing platform to form a government data set, which is then encrypted.

[0050] The government Internet of Things devices include self-service devices, including smart guide robots and self-service terminals of government service agencies, and encrypt government data sets of the government Internet of Things through encryption algorithms and key generation, storage, distribution and update mechanisms.

[0051] Intelligent guide robot: Provides navigation and consulting services in places such as government halls, guiding people to quickly find the service windows they need. The intelligent guide robot has a rich business knowledge base built in, covering the various service contents, handling procedures, required materials and other information provided by the government hall. It also has emotion recognition function and can sense the emotional changes of the public. It provides key information such as government needs, response distribution and user behavior patterns.

[0052] Self-service terminal: provides self-service functions for various government services, such as inquiry, payment, printing, etc. It provides key information such as government needs, response distribution, and user behavior patterns.

[0053] The real-time data items collected include but are not limited to government applicant information, government information type, service handling volume, handling time, handling results, public satisfaction evaluation, complaint and suggestion data, etc. The government data sets of the government IoT are encrypted through encryption algorithms and key generation, storage, distribution and update mechanisms. The real-time and integrity of the data are ensured, and only the minimum amount of data required to implement the business is collected and processed, reducing the potential risk of data leakage.

[0054] Using IoT technology, data is transmitted to the cloud or local data center through wireless sensor networks. The encryption and integrity of data transmission are ensured to prevent data from being tampered with or stolen during transmission. Edge computing technology can also be used to process part of the data, reduce transmission delays and improve system response speed.

[0055] Then, the maximum possible change of each data item and the query result corresponding to the data item in the government data set is calculated by the differential privacy algorithm, and the sensitivity Δf of each data item and the query result corresponding to the data item in the government data set is evaluated;

[0056] In the government IoT, a large amount of real-time data needs to be collected, processed and published. This data may contain sensitive information such as the user's basic government information, government needs, and device status. Through the differential privacy algorithm, the data can be processed before it is released, and appropriate noise can be added to prevent attackers from inferring specific user information through data analysis, thereby protecting government privacy. During data analysis, the original data is processed through the differential privacy algorithm to ensure that the analysis results will not change significantly due to the presence or absence of a specific data, thereby protecting the privacy information in the data. This helps government enterprises make full use of data resources for optimized decision-making and predictive analysis while ensuring government privacy.

[0057] For non-numeric data, the square root of the sum of the squares of the changes in the query results corresponding to the data items caused by the changes in the data items is calculated to obtain the sensitivity Δf of each data item in the government data set and the query results corresponding to the data items. Scores are assigned to different query results and sorted according to the scores. The query results with higher scores have higher sensitivity Δf.

[0058] For numerical data, the maximum absolute change in the query result corresponding to the data item when the data item changes is calculated, Δf = max_{x,x'∈D}|f(x)-f(x')|, where D is the data set, x and x' are two adjacent data items in the data set, f is the query function, and Δf is the maximum change in the query result;

[0059] The higher the sensitivity Δf of a data item or the query result corresponding to the data item, the greater the threat to privacy posed by the data item or query result, and more random noise needs to be added. Sensitivity determines the amount of noise that needs to be added, and highly sensitive data items will require more noise to cover up their changes. The addition of noise should ensure the strictness of differential privacy, that is, the change of any single data item will not significantly affect the distribution of query results.

[0060] Non-numeric data: For non-numeric data, sensitivity assessment usually involves calculating the square root of the sum of the squares of the changes in query results when the data item changes. This method can quantify the impact of the change in the data item on the query result, thereby assessing its sensitivity.

[0061] Numerical data: For numerical data (such as the amount of government service transactions, the duration of transactions, etc.), sensitivity assessment usually calculates the maximum absolute change in the query results when the data item changes. This is because a small change in numerical data may have a significant impact on the query results, so a more rigorous sensitivity assessment method is required.

[0062] Through sensitivity assessment, we can identify which data items and query results pose a greater threat to privacy, so that more random noise can be added to them to protect privacy.

[0063] Next, according to the sensitivity Δf and privacy budget ε of each data item and the query result corresponding to the data item in the government data set, random noise is added to the data item and the query result corresponding to the data item;

[0064] For numerical data, such as the amount of government service processing and processing time, noise that follows the Laplace distribution Lap(Δf / ε) is added to the data items and the query results of the data items; Δf is the sensitivity of the data item or query result, and ε is the privacy budget parameter, which determines the degree of privacy protection and the amount of noise added. The smaller the privacy budget, the more noise is added, and the higher the degree of privacy protection.

[0065] For non-numerical data, such as processing results, public satisfaction evaluation, complaint and suggestion data, etc., the scores of different query results are evaluated and sorted through a scoring function. The probability distribution of the output query results is adjusted according to the sorting results and the privacy budget ε to reduce the certainty of specific results.

[0066] According to the sensitivity Δf and privacy budget ε calculated by the differential privacy algorithm, the scale parameter of the noise is determined to be Δf / ε; a random number generator is used to generate random noise that follows the Laplace distribution, and a noise value is randomly extracted; the generated noise value is added to the government data set or query results. For numerical data, the noise value is directly added to the original data; for non-numerical data, noise is indirectly added by adjusting the probability distribution of the query results.

[0067] Taking adding random noise to numerical data as an example, first determine the privacy budget ε: for example, set ε = 1, and then calculate the sensitivity Δf: according to the characteristics of the query function and the data set, calculate the sensitivity Δf. Determine the noise parameter: use the Laplace distribution, whose scale parameter is related to the sensitivity Δf and the privacy budget ε, which is Δf / ε. Generate noise: randomly extract a noise value from the Laplace distribution. Add noise: add the noise value to the query result. Publish or analyze the results: publish the query results after adding noise to relevant personnel or conduct further analysis. For queries on numerical data, such as count queries, for such queries, the sensitivity Δf is usually 1, so Lap(1 / ε) can be directly used as noise to add.

[0068] Then, a machine learning algorithm is used to analyze the government service efficiency, government anomaly monitoring, government demand forecasting, and government complaint risk forecasting on the government data set with random noise added.

[0069] Government service efficiency analysis: By analyzing historical data, including output indicators such as service volume, service quality, and public satisfaction, we can identify the changing trends of government demand and response, provide a basis for future government planning and scheduling, analyze the design and implementation of government service processes, including the degree of process simplification, collaborative efficiency, and information sharing, evaluate the effectiveness and level of government services, and then optimize government service processes, improve service quality, and strengthen information construction;

[0070] Government affairs abnormality monitoring and analysis: Real-time monitoring of abnormal events in government affairs, including security incident monitoring, business processing abnormalities, etc. Specifically, real-time monitoring of the security status of government affairs systems, including intrusion detection, virus protection, data encryption, etc. Once a security threat or attack is discovered, immediate response measures are taken, such as blocking the source of attack, repairing security vulnerabilities, strengthening security protection, etc., and key links in the government affairs business process are monitored to ensure the continuity and stability of business operations. For example, in the application, approval, and processing of government services, monitor whether there are timeouts, unprocessed, and error handling, and promptly remind relevant personnel to handle them. Timely discovery and resolution of problems ensures the safe and stable operation of government affairs;

[0071] Government demand forecasting and analysis: Based on historical data and current trends, forecast future government demand to provide guidance for the addition and allocation of government items; based on the results of trend analysis, forecast new demands and new problems that may arise in the future to provide a basis for the addition of government items. Priority should be given to adding government service projects that are closely related to public needs to improve the pertinence and effectiveness of government services. Government demand forecasting can provide a scientific basis for government decision-making and help decision-makers better understand the focus and direction of future government work. By predicting future government demand, the government can allocate human, material and financial resources more reasonably to ensure the smooth progress of government work. Understanding and predicting the needs of the masses will help the government take measures in advance, improve service processes, improve service quality, and meet the expectations of the masses.

[0072] Government complaint risk prediction: Build a government complaint risk scoring model through government data and communication data to improve the accuracy and efficiency of government complaint risk prediction.

[0073] The government affairs data set is divided into government affairs training set, government affairs verification set and government affairs test set. The government affairs training set is used to train the model, the government affairs verification set is used to adjust the model parameters, and the government affairs test set is used to evaluate the model performance. The model is trained using the government affairs training set, and the model parameters are adjusted by minimizing the prediction error. During the training process, cross-validation and other techniques are used to prevent overfitting. The trained model is evaluated using the government affairs test set, and indicators such as prediction accuracy and recall rate are calculated to measure the performance of the model. The model parameters are adjusted according to the evaluation results, or other models are selected for trial. Among them, a linear regression model is used to predict continuous numerical data, such as government affairs response prediction, and a decision tree is used for classification and regression tasks, such as government affairs fault diagnosis. A neural network model is used to train nonlinear problems, such as government affairs consumption optimization.

[0074] The analysis results are then provided as data services through the API interface, using a strong identity authentication mechanism, and setting access control rules based on user roles and permissions. The access rights to each data item in the government data set and the query results corresponding to the data item are set through the access control list ACL;

[0075] The strong authentication mechanism is a multi-factor authentication mechanism, including username / password, dynamic password, biometrics, etc. The access control rules set according to user roles and permissions are specifically divided into different roles, and a set of permissions are defined for each role. The specific user roles are defined as follows: Administrator: has full access to government data sets, including the addition, deletion, modification and query of data items and the management of other user permissions. Data analyst: Responsible for data analysis, has read permissions to specific data items and their query results, but usually does not have permissions to modify or delete data. Operation and maintenance personnel: Responsible for the daily operation and maintenance of data systems, may need to access system logs or monitoring data, but are not directly involved in the processing of business data. Ordinary users: Depending on business needs, may only be able to access limited data items or query results.

[0076] Access rules are set according to role permissions: Rule 1: Administrators can access all data items and perform all operations, including reading, writing, modifying, and deleting; Rule 2: Data analysts can read specific data items, including power statistics and query results in response to prediction results, but cannot modify or delete data; Rule 3: Operation and maintenance personnel can access system logs and monitoring data, but cannot access business data; Rule 4: Ordinary users are authorized to access limited data items or query results based on their business needs.

[0077] The analysis results are presented to stakeholders such as government service providers and government agencies in a visual form (such as charts and dashboards). Data services are provided through API interfaces to support third-party developers to build applications based on government IoT data. Through the API interface, users can easily access data and analysis results from anywhere and at any time without directly accessing the original database or data analysis system. Improves the accessibility and convenience of data. The API interface allows seamless data exchange and integration between different systems, applications and platforms. The analysis results can be integrated into existing business processes, applications or third-party services to achieve comprehensive integration and efficient use of data. It can respond to market changes and customer needs more quickly. When business needs change, the data services provided by the API interface can be quickly adjusted without large-scale modifications to the entire system. The API interface makes it easier to share data between teams, departments or organizations. When providing data services through the API interface, strict security measures such as authentication, authorization and encryption can be implemented to protect the security and privacy of data.

[0078] During the release process, ensure that no information that may endanger personal privacy or organizational security is leaked. It can effectively balance the needs of data privacy protection and data utilization, and provide strong support for the development of smart government affairs.

[0079] Finally, the above analysis results and query results are evaluated, and the sensitivity Δf, privacy budget ε, and the degree of random noise addition are dynamically adjusted according to the evaluation results to adapt to different privacy protection requirements and data changes, ensuring that the practicality and accuracy of the data are maximized while protecting privacy.

[0080] Sensitivity adjustment: Re-evaluate the sensitivity of data items and query results based on changes in data and business requirements, and adjust their values ​​accordingly.

[0081] Privacy budget adjustment: According to the privacy protection requirements and the accuracy requirements of data analysis, adjust the value of the privacy budget ε. The larger the privacy budget, the less noise is added, but the lower the privacy protection; the smaller the privacy budget, the more noise is added, but the higher the privacy protection.

[0082] Adjust the degree of noise addition: Recalculate and add random noise based on the adjusted sensitivity and privacy budget to ensure data privacy and analysis accuracy.

[0083] Regularly evaluate the effectiveness of encryption and privacy protection to check whether the degree of privacy protection meets the preset goals and whether the data availability meets business needs. According to the evaluation results, adjust the privacy budget, sensitivity analysis and noise addition strategy in a timely manner to optimize the balance between privacy protection and data utilization.

[0084] The present invention also provides an IoT service data encryption and privacy protection device, comprising:

[0085] The data collection and encryption processing unit uses the Internet of Things technology to collect real-time data items from government Internet of Things devices using a wireless sensor network and transmits them to the cloud computing platform to form a government data set. The government data set is encrypted and then output to the data sensitivity assessment unit;

[0086] The data sensitivity evaluation unit calculates the maximum possible change of each data item in the government data set and the query result corresponding to the data item through a differential privacy algorithm, evaluates the sensitivity Δf of each data item in the government data set and the query result corresponding to the data item, and the sensitivity Δf evaluation result is transferred to the data noise adding unit;

[0087] A data noise adding unit, which adds random noise to the data item and the query result corresponding to the data item according to the sensitivity Δf and the privacy budget ε of each data item in the government data set and the query result corresponding to the data item, and outputs the government data set after adding the random noise to the government data set analysis unit for analysis;

[0088] The government data set analysis unit uses a machine learning algorithm to perform government service efficiency analysis, government anomaly monitoring analysis, government demand forecasting analysis, and government complaint risk forecasting analysis on the government data set after random noise is added, and the analysis results are output to the interaction and authority control unit and the dynamic adjustment unit;

[0089] The interaction and permission control unit provides data services through the API interface with the analysis results, adopts a strong identity authentication mechanism, and sets access control rules according to user roles and permissions. It sets access rights to each data item in the government data set and the query results corresponding to the data item through the access control list ACL, and the query results are fed back to the dynamic adjustment unit;

[0090] The dynamic adjustment unit evaluates the analysis results in the government data set analysis unit and the query results in the interaction and permission control unit, dynamically adjusts the sensitivity Δf, privacy budget ε and the degree of random noise addition according to the evaluation results, and feeds back to the data sensitivity evaluation unit and the data noise addition unit.

[0091] By integrating multiple functional units, comprehensive protection, security analysis and dynamic adjustment of government IoT data are achieved, providing strong support for the safe operation and efficient management of smart government affairs.

[0092] The present invention can be applied to the following scenarios:

[0093] Government response prediction: When predicting government response, it is necessary to integrate multiple data sources, such as historical response time, government processing time, user behavior data, etc. By processing these data with the method of the present invention, the accuracy and reliability of response prediction can be improved while protecting government privacy.

[0094] Government Complaint Risk Prediction: Government complaint risk prediction is an important means to reduce the government complaint rate. By integrating government data and communication data, the method of the present invention can be used to construct a government complaint risk scoring model.

[0095] Government service energy management and optimization: In the field of government service energy management and optimization, the method of the present invention can be used to protect government privacy while achieving efficient allocation and utilization of government service energy.

[0096] Any example of the present invention can be used as an independent technical solution or combined with other examples. All patents and publications mentioned in the specification of the present invention indicate that these are public technologies in the field and can be used in the present invention. All patents and publications cited here are also listed in the references, just as each publication is specifically cited separately. The present invention here can be implemented in the absence of any element or elements, one limitation or multiple limitations, and such limitations are not specifically stated here. The terms and expressions used here are descriptive methods, but are not limited by them. There is no intention to indicate that these terms and explanations described in this book exclude any equivalent features, but it can be known that any appropriate changes or modifications can be made within the scope of the present invention and the claims. It can be understood that the embodiments described in the present invention are embodiments and features in some embodiments, and any person skilled in the art can make some changes and variations based on the essence of the present invention, and these changes and variations are also considered to belong to the scope of the present invention and the scope limited by the independent claims and the appended claims.

Claims

1. A method for encrypting and protecting privacy of IoT service data, characterized in that: The specific steps include: S1. Through the Internet of Things technology, wireless sensor networks are used to collect real-time data items from government Internet of Things devices and transmit them to the cloud computing platform to form a government data set, and the government data set is encrypted; S2. Calculate the maximum possible change of each data item and the query result corresponding to the data item in the government data set through the differential privacy algorithm, and evaluate the sensitivity Δf of each data item and the query result corresponding to the data item in the government data set; S3, adding random noise to the data item and the query result corresponding to the data item according to the sensitivity Δf and the privacy budget ε of each data item and the query result corresponding to the data item in the government data set; S4. Use machine learning algorithms to perform government service efficiency analysis, government anomaly monitoring analysis, government demand forecasting analysis, and government complaint risk forecasting analysis on government data sets with random noise added; S5. Provide data services through API interfaces with analysis results, adopt strong identity authentication mechanisms, set access control rules based on user roles and permissions, and set access rights to each data item in the government data set and the query results corresponding to the data item through the access control list ACL; S6. Evaluate the analysis results and query results in S4 and S5, and dynamically adjust the sensitivity Δf, privacy budget ε, and the degree of random noise addition according to the evaluation results.

2. According to claim 1, a method for encrypting and protecting privacy of Internet of Things business data is characterized in that: The S1 specifically includes: the government Internet of Things devices include self-service devices, including smart guide robots and self-service terminals of government service agencies, which collect real-time data items including but not limited to government applicant information, government information type, service processing volume, processing time, processing results, public satisfaction evaluation, complaint and suggestion data, etc., and encrypt the government data set of the government Internet of Things through encryption algorithms and key generation, storage, distribution and update mechanisms.

3. The method for encrypting and protecting privacy of IoT service data according to claim 1, characterized in that: The sensitivity Δf of evaluating each data item in the government data set and the query result corresponding to the data item in S2 is specifically as follows: for non-numeric data, the square root of the sum of the squares of the changes in the query result corresponding to the data item caused by the change in the data item is calculated to obtain the sensitivity Δf of each data item in the government data set and the query result corresponding to the data item, and scores are assigned to different query results, and they are sorted according to the scores. The query result with a higher score has a higher sensitivity Δf; For numerical data, the maximum absolute change in the query result corresponding to the data item when the data item changes is calculated, Δf = max_{x,x'∈D}|f(x)-f(x')|, where D is the data set, x and x' are two adjacent data items in the data set, f is the query function, and Δf is the maximum change in the query result; The higher the sensitivity Δf of a data item or the query result corresponding to the data item, the more random noise needs to be added.

4. The method for encrypting and protecting privacy of IoT service data according to claim 1, characterized in that: In S3, random noise is added to the data items and the query results corresponding to the data items as follows: for numerical data, noise that obeys the Laplace distribution Lap(Δf / ε) is added to the query results corresponding to the data items and the data items; for non-numerical data, the scores of different query results are evaluated and sorted by a scoring function, and the probability distribution of the output query results is adjusted according to the sorting results and the privacy budget ε.

5. The method for encrypting and protecting privacy of IoT service data according to claim 4, characterized in that: Adding random noise to numerical data in S3 also includes: S31. According to the sensitivity Δf and the privacy budget ε calculated by the differential privacy algorithm, the scale parameter of the noise is determined to be Δf / ε. S32, using a random number generator to generate random noise that obeys Laplace distribution, and randomly extracting a noise value; S33. Add the generated noise value to the government data set or query result.

6. The method for encrypting and protecting privacy of IoT service data according to claim 1, characterized in that: The S4 specifically includes: Government service efficiency analysis: By analyzing historical data, we can identify the changing trends of government service demand and response efficiency, and provide a basis for future government planning and scheduling; Government affairs abnormality monitoring: Real-time monitoring of abnormal events in government affairs, including security event monitoring, business processing abnormalities, etc., timely discovery and resolution of problems to ensure the safe and stable operation of government affairs; Government demand forecasting: Based on historical data and current trends, forecast future government demand and provide guidance for government item addition and allocation; Government complaint risk prediction: Build a government complaint risk scoring model through government data and communication data to improve the accuracy and efficiency of government complaint risk prediction.

7. The method for encrypting and protecting privacy of IoT service data according to claim 1, characterized in that: The S4 specifically includes: dividing the government data set into a government training set, a government verification set and a government test set, the government training set is used to train the model, the government verification set is used to adjust the model parameters, and the government test set is used to evaluate the model performance. The model is trained using the government training set, and the model parameters are adjusted by minimizing the prediction error. During the training process, cross-validation and other techniques are used to prevent overfitting. The trained model is evaluated using the government test set, and indicators such as prediction accuracy and recall rate are calculated to measure the performance of the model. The model parameters are adjusted according to the evaluation results or other models are selected for trial. Among them, a linear regression model is used to predict continuous numerical data, such as government response prediction, a decision tree is used to perform classification and regression tasks, such as government fault diagnosis, and a neural network model is used to train nonlinear problems, such as government consumption optimization.

8. The method for encrypting and protecting privacy of IoT service data according to claim 1, characterized in that: The S6 specifically includes: Sensitivity adjustment: Re-evaluate the sensitivity of data items and query results based on changes in data and business requirements, and adjust their values ​​accordingly. Privacy budget adjustment: According to the privacy protection requirements and the accuracy requirements of data analysis, adjust the value of the privacy budget ε. The larger the privacy budget, the less noise is added, but the lower the privacy protection; the smaller the privacy budget, the more noise is added, but the higher the privacy protection. Adjust the degree of noise addition: Recalculate and add random noise based on the adjusted sensitivity and privacy budget to ensure data privacy and analysis accuracy.

9. The method for encrypting and protecting privacy of IoT service data according to claim 1, characterized in that: The S5 specifically includes: the strong identity authentication mechanism is a multi-factor identity authentication mechanism, including username / password, dynamic password, biometrics, etc. The access control rules set according to user roles and permissions are specifically divided into different roles, and a set of permissions are defined for each role, and access rules are set according to role permissions: Rule 1: Administrators can access all data items and perform all operations, including reading, writing, modifying, and deleting; Rule 2: Data analysts can read specific data items, including power statistics data and query results in response to prediction results, but cannot modify or delete data; Rule 3: Operation and maintenance personnel can access system logs and monitoring data, but cannot access business data; Rule 4: Ordinary users are authorized to access limited data items or query results based on their business needs.

10. An Internet of Things service data encryption and privacy protection device, using the Internet of Things service data encryption and privacy protection method as described in any one of claims 1 to 9, characterized in that: include: The data collection and encryption processing unit uses the Internet of Things technology to collect real-time data items from government Internet of Things devices using a wireless sensor network and transmits them to the cloud computing platform to form a government data set. The government data set is encrypted and then output to the data sensitivity assessment unit; The data sensitivity evaluation unit calculates the maximum possible change of each data item in the government data set and the query result corresponding to the data item through a differential privacy algorithm, evaluates the sensitivity Δf of each data item in the government data set and the query result corresponding to the data item, and the sensitivity Δf evaluation result is transferred to the data noise adding unit; A data noise adding unit, which adds random noise to the data item and the query result corresponding to the data item according to the sensitivity Δf and the privacy budget ε of each data item in the government data set and the query result corresponding to the data item, and outputs the government data set after adding the random noise to the government data set analysis unit for analysis; The government data set analysis unit uses a machine learning algorithm to perform government service efficiency analysis, government anomaly monitoring analysis, government demand forecasting analysis, and government complaint risk forecasting analysis on the government data set after random noise is added, and the analysis results are output to the interaction and authority control unit and the dynamic adjustment unit; The interaction and permission control unit provides data services through the API interface with the analysis results, adopts a strong identity authentication mechanism, and sets access control rules according to user roles and permissions. It sets access rights to each data item in the government data set and the query results corresponding to the data item through the access control list ACL, and the query results are fed back to the dynamic adjustment unit; The dynamic adjustment unit evaluates the analysis results in the government data set analysis unit and the query results in the interaction and permission control unit, dynamically adjusts the sensitivity Δf, privacy budget ε and the degree of random noise addition according to the evaluation results, and feeds back to the data sensitivity evaluation unit and the data noise addition unit.

Images