Trusted data management and control system and method

Abstract

The invention relates to the technical field of data security and privacy protection, in particular to a trusted data management and control system and method. The data input unit is used for collecting multi-source user transaction data, identifying preset sensitive data in the collection process, performing de-identification processing on the sensitive data, encrypting the collected transaction data in combination with homomorphic encryption and symmetric encryption algorithms, storing the encrypted transaction data, and storing the stored transaction data; the access control unit stores the transaction data, obtains the transaction data needing to be analyzed from the stored transaction data according to the user input requirement of the access control unit, decrypts the transaction data before obtaining the transaction data, and optimizes the transaction data by using a differential privacy method in the analysis process. According to the trusted data management and control system and method, when a user processes different query tasks, the privacy budget can be flexibly and accurately adjusted, and the accuracy of privacy protection and the reliability of query results can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data security and privacy protection, and in particular to a trusted data management and control system and method. Background Art

[0002] With the rapid development of big data and artificial intelligence technologies, the financial industry is increasingly dependent on data, and the collection, analysis and processing of transaction data have become key tasks. However, transaction data in the financial industry usually contains a large amount of sensitive information, such as personal accounts, transaction amounts, transaction times, etc. Once this information is leaked, it may lead to serious risks such as privacy leakage, identity theft and financial fraud.

[0003] Privacy budget has become a key concept in protecting privacy and the security of sensitive information. Privacy budget is used to quantify the degree of privacy protection in data analysis, which is usually achieved through differential privacy methods, that is, by adding noise to the data query results to avoid leaking sensitive information of a single data point. In financial transaction data processing, the setting of privacy budget directly affects the strength of privacy protection and the accuracy of query results.

[0004] The privacy budget calculation in the existing technology often cannot be adjusted according to the needs of different queries, which easily leads to inefficient privacy protection and data analysis. In traditional methods, the privacy budget may be too high, resulting in excessive noise addition, reducing the accuracy of query results; or the privacy budget is too low, privacy protection is insufficient, and there is a risk of leakage. Therefore, a trusted data management system and method are provided. Summary of the invention

[0005] The purpose of the present invention is to provide a trusted data management system and method to solve the problem that the privacy budget calculation proposed in the above background technology often cannot be adjusted according to the needs of different queries, which easily leads to low efficiency of privacy protection and data analysis.

[0006] To achieve the above objectives, on the one hand, the present invention aims to provide a trusted data management and control system, including: A data input unit, the data input unit is used to collect multi-source user transaction data, identify pre-set sensitive data during the collection process, de-identify the sensitive data, and transmit the processed transaction data to the encryption protection unit; An encryption protection unit, which is used to encrypt the collected transaction data by combining homomorphic encryption and symmetric encryption algorithms, and after encryption is completed, transmit the encrypted transaction data to the storage analysis unit; A storage and analysis unit, the storage and analysis unit is used to store the encrypted transaction data, obtain the transaction data to be analyzed from the stored transaction data according to the requirements input by the user of the access control unit, decrypt the transaction data before obtaining the transaction data, and use the differential privacy method for optimization during the analysis process; An access control unit, the access control unit being used to manage access rights to the stored transaction data according to the rights and needs of the user; The audit monitoring unit is used to monitor and record in real time the access, operation and change activities of all transaction data in the system, and can also track user behavior, detect abnormal activities, and generate audit reports.

[0007] As a further improvement of the present technical solution, the pre-set sensitive data includes user personal information, user IP address, transaction device ID and transaction amount.

[0008] As a further improvement of the technical solution, in the encryption protection unit, the collected transaction data is encrypted by combining homomorphic encryption and symmetric encryption algorithms, and the specific steps are as follows: Use homomorphic encryption for sensitive data: ; in, It is the sensitive data after homomorphic encryption; The original sensitive data; It is a homomorphic encryption operation; Use symmetric encryption for non-sensitive data: ; in, The ciphertext is the symmetrical encryption of non-sensitive data; is non-sensitive data; It is a symmetric encryption algorithm.

[0009] As a further improvement of the present technical solution, the encryption protection unit also includes a key management module, which can rotate the key using a hybrid key rotation strategy in a trusted execution environment. At the same time, when the key expires or is threatened with leakage, the key management module triggers the key update process, and the updated key will immediately replace the old key. If the key is leaked or cracked, the key management module can revoke the use of the key.

[0010] As a further improvement of the technical solution, the specific steps of rotating the key by the key management module in the trusted execution environment using the hybrid key rotation strategy are as follows: Based on the audit monitoring unit, monitor the usage frequency and real-time risks of keys; Rotation conditions based on frequency of use: ; in, is the key; is the number of times the key is used; is the usage count threshold; The usage time of the key; is the usage duration threshold; Risk-based rotation criteria: ; in, The risk value is calculated based on the real-time risk of transaction data; The risk thresholds to be set; Trigger conditions for the hybrid rotation strategy: ; If any of the conditions is met, key rotation is triggered, and then the old key is backed up and a new key is generated, the new key is updated and the old key is destroyed. At the same time, all key rotation operations are recorded and transmitted to the audit monitoring unit.

[0011] As a further improvement of the technical solution, in the storage and analysis unit, the specific steps of decrypting the transaction data before acquiring the transaction data are as follows: Use batch operations to aggregate multiple encrypted sensitive data: ; in, is the total sensitive data that has been homomorphically encrypted; For each sensitive data Homomorphic encryption result of ; The amount of sensitive data; ; Represents a batch homomorphic addition operation; Homomorphically decrypt the encrypted total sensitive data: ; in, The total sensitive data for homomorphic encryption Decryption is performed, and the decrypted result is the total transaction amount in plain text; is the total sensitive data after decryption; Symmetrically decrypt sensitive data: ; in, For non-sensitive data after symmetrical encryption Decrypt and obtain the plaintext data; It is the ciphertext data after symmetrical encryption; The decrypted sensitive data.

[0012] As a further improvement of the technical solution, in the storage analysis unit, a differential privacy method is used for optimization during the analysis process, and the optimization is as follows: First, determine the query operation entered by the user and adjust the query sensitivity according to the type of transaction data query task. ; Then identify the importance of the query task, based on the importance of the query task and the sensitivity of the query and dynamically allocate privacy budgets based on expected privacy needs ; Then, the noise type is selected according to the type of query task and the characteristics of transaction data, and the noise is added to the query process, so as to perform the query and output the result after adding noise.

[0013] As a further improvement of this technical solution, the sensitivity Adjustment and privacy budget The choices are as follows: ; in, To query the sensitivity of the task; Transaction data points Query results; Transaction data points Query results; is the adaptive adjustment factor; ; in, For the original The privacy budget assigned to each query task; is a constant coefficient; is the importance weight of the query task; is the dynamic adjustment factor of the query task; is the size of the transaction dataset; Sensitivity correction factor for the query type.

[0014] As a further improvement of this technical solution, the upper and lower limit constraints of the privacy budget are added to the privacy budget formula for optimization. The optimization is as follows: ; in, For the The privacy budget assigned to each query task; is the minimum value of the privacy budget; is the maximum value of the privacy budget.

[0015] On the other hand, the present invention provides a trusted data management method, based on the above-mentioned trusted data management system, comprising the following steps: S1. Collect multi-source user transaction data, identify pre-set sensitive data during the collection process, de-identify the sensitive data, and then encrypt the collected transaction data using a combination of homomorphic encryption and symmetric encryption algorithms; S2. Store the encrypted transaction data and manage access rights to the stored transaction data according to the user's authority and needs; S3. Obtain the transaction data that needs to be analyzed from the stored transaction data according to the user's needs, decrypt the transaction data before obtaining it, and use the differential privacy method to optimize it during the analysis process; S4. Monitor and record all access, operation and change activities of transaction data in the system in real time, track user behavior, detect abnormal activities, and generate audit reports.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. In this trusted data management and control system and method, a balance between efficient computing and high security is achieved by combining homomorphic encryption and symmetric encryption. Specifically, homomorphic encryption is used to protect data privacy, encrypt data during storage, transmission and calculation, and symmetric encryption is used to accelerate the protection and calculation process of encrypted data, while reducing the computational overhead of homomorphic encryption.

[0017] 2. In the trusted data management system and method, users can flexibly and accurately adjust the privacy budget when processing different query tasks, which helps to improve the accuracy of privacy protection and the reliability of query results. At the same time, by preventing the privacy budget from being too high or too low, the potential impact of the system on the query results is reduced, thereby improving the stability and security of the overall data processing process. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is the overall flow chart of the present invention; Figure 2 A flow chart of the overall method of the present invention; The meaning of each number in the figure is: 1. Data input unit; 2. Encryption protection unit; 21. Key management module; 3. Storage analysis unit; 4. Access control unit; 5. Audit monitoring unit. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] Example 1: Please refer to Figure 1 As shown, a trusted data management and control system is provided, including a data input unit 1, an encryption protection unit 2, a storage analysis unit 3, an access control unit 4 and an audit monitoring unit 5; The data input unit 1 is used to collect multi-source user transaction data, identify pre-set sensitive data during the collection process, de-identify the sensitive data, and transmit the processed transaction data to the encryption protection unit 2; The pre-set sensitive data includes user personal information, user IP address, transaction device ID, and transaction amount; The encryption protection unit 2 is used to encrypt the collected transaction data by combining homomorphic encryption and symmetric encryption algorithms. After encryption is completed, the encrypted transaction data is transmitted to the storage analysis unit 3; In encryption protection unit 2, the collected transaction data is encrypted by combining homomorphic encryption and symmetric encryption algorithms. The specific steps are as follows: Use homomorphic encryption for sensitive data: ; in, It is the sensitive data after homomorphic encryption. This is the encrypted transaction data, which can be directly calculated in the encrypted state; The original sensitive data (such as transaction amount) is the plaintext data before homomorphic encryption; It is a homomorphic encryption operation; Use symmetric encryption for non-sensitive data: ; in, The ciphertext is the symmetrical encryption of non-sensitive data; is non-sensitive data, such as a user's identifier or other data that poses no privacy risk; is a symmetric encryption algorithm. is the encryption key. The symmetric encryption here is used to encrypt non-sensitive information. Only sensitive data (such as transaction amount) is homomorphically encrypted, and all data (such as transaction ID, timestamp) is not encrypted repeatedly. For non-sensitive data, the plain text is directly stored or a simpler encryption method (such as symmetric encryption AES) is used. In this way, we reduce the computational overhead and maintain the advantages of homomorphic encryption. Although homomorphic encryption can perform calculations in an encrypted state, due to its high computational complexity, especially fully homomorphic encryption, performance becomes a bottleneck when processing large-scale data; symmetric encryption is more efficient when encrypting and decrypting, but it cannot directly support computing operations on encrypted data.

[0021] By combining homomorphic encryption and symmetric encryption, a balance between efficient computing and high security is achieved. The specific strategy is: use homomorphic encryption to protect data privacy and encrypt data during storage, transmission and calculation; use symmetric encryption to accelerate the protection and calculation process of encrypted data, while reducing the computational overhead of homomorphic encryption.

[0022] The encryption protection unit 2 also includes a key management module 21, which can rotate the key using a hybrid key rotation strategy in a trusted execution environment. At the same time, when the key expires or is threatened with leakage, the key management module 21 triggers the key update process, and the updated key will immediately replace the old key. If the key is leaked or cracked, the key management module 21 can revoke the use of the key, and the revocation operation will ensure that the key marked as invalid is no longer used in the system; The specific steps of the key management module 21 rotating the key using the hybrid key rotation strategy in the trusted execution environment are as follows: Based on the audit monitoring unit 5, monitor the usage frequency and real-time risk of the key; Rotation conditions based on frequency of use: ; in, is the key; The number of times the key is used. Each time the key is used for encryption or decryption, the counter increases. This helps the system understand if the key is overused; The usage count threshold. When the number of times a key is used reaches or exceeds this threshold, it means that the key has been used frequently and needs to be rotated to prevent the risk of leakage. The usage time of the key; The usage time threshold indicates the maximum life cycle of the key. When the usage time of the key exceeds this threshold, rotation is triggered, even if it has not reached the usage threshold; That is, when the number of times or time of use exceeds the threshold, the system triggers the key rotation operation; Risk-based rotation criteria: ; in, The risk value is calculated based on the real-time risk of transaction data (such as intrusion detection, abnormal access, etc.). The risk level is usually calculated based on factors such as the security status of the system, access patterns, and abnormal behaviors detected. For example, abnormal access, suspicious activities in system logs, and intrusion detection system (IDS) reports may affect risk assessment; The risk threshold is set. When the risk level detected by the system exceeds this threshold, it means that the system is in a high-risk state and may be threatened by attacks or leaks, so key rotation is required immediately; That is, when the risk level exceeds the set risk threshold, the system will immediately trigger the key rotation operation; Trigger conditions for the hybrid rotation strategy: ; If any of the conditions is met, key rotation is triggered, and then the old key is backed up and a new key is generated, the new key is updated and the old key is destroyed. At the same time, all key rotation operations are recorded and transmitted to the audit monitoring unit 5, including the steps of backing up, generating a new key, updating the key and destroying the old key; For example, in a financial transaction system, the transaction key may quickly reach the usage threshold due to frequent transactions. At this time, the system first triggers key rotation according to the rotation strategy based on usage frequency. At the same time, if the system detects transaction anomalies (such as frequent login failures or suspected DDoS attacks), the rotation frequency is increased according to the risk-based rotation strategy to reduce security risks; By combining the key rotation based on the frequency of use and the intelligent risk-based key rotation method, the new hybrid rotation strategy can dynamically adjust the rotation strategy based on real-time risk assessment while ensuring the high frequency of key use. This strategy not only ensures the security of the system, but also avoids overly frequent rotation, improving the efficiency and flexibility of the system.

[0023] The storage and analysis unit 3 is used to store the encrypted transaction data, and obtain the transaction data to be analyzed from the stored transaction data according to the requirements of the user input of the access control unit 4, decrypt the transaction data before obtaining the transaction data, and use the differential privacy method for optimization during the analysis process; In the storage and analysis unit 3, the specific steps of decrypting the transaction data before acquiring the transaction data are as follows: Use batch operations to aggregate multiple encrypted sensitive data: ; in, is the total sensitive data that has been homomorphically encrypted; For each sensitive data Homomorphic encryption result of ; The amount of sensitive data; ; Represents batch homomorphic addition operations, which can complete multiple addition operations in one step; Batch processing allows you to process multiple encrypted data points at once, rather than processing each individual data item one by one. This not only reduces repeated calculations, but also increases the overall processing speed. , …, , aggregation can be performed in the same round of calculation. For example, for application scenarios such as financial transactions that require frequent processing of large amounts of sensitive information, batch processing can better support complex analytical tasks, such as total amounts or other summary indicators, without directly accessing or exposing each individual transaction record.

[0024] Homomorphically decrypt the encrypted total sensitive data: ; in, The total sensitive data for homomorphic encryption Decryption is performed, and the decrypted result is the total transaction amount in plain text; is the total sensitive data after decryption; Symmetrically decrypt sensitive data: ; in, For non-sensitive data after symmetrical encryption Decrypt and obtain the plaintext data; It is the ciphertext data after symmetrical encryption; The decrypted sensitive data; With layered decryption, aggregate data is partially decrypted first, and then further decrypted as needed. For example, in financial transaction data processing, we first decrypt only the aggregate data of the total amount, and then perform the next operation based on this data.

[0025] In the storage analysis unit 3, the differential privacy method is used for optimization during the analysis process. Differential privacy is to add noise to the data to ensure that individual information is not exposed during data analysis, which is suitable for sensitive data sets. The optimization is as follows: First, determine the query operation entered by the user (such as average, count, sum, etc.), and adjust the query sensitivity according to the type of transaction data query task ; Then identify the importance of the query task, based on the importance of the query task and the sensitivity of the query and dynamically allocate privacy budgets based on expected privacy needs ; Then, the noise type is selected according to the type of query task and the characteristics of transaction data, and the noise is added to the query process, so as to perform the query and output the result after adding noise.

[0026] The core purpose of differential privacy is to protect the information of individuals in the data and prevent the leakage of sensitive data. By adding noise (such as Laplace noise or Gaussian noise), even if an attacker obtains the analysis results, he cannot accurately infer the data of a specific individual. , which can prevent the influence of a single data point from being exposed in the query results, thereby protecting privacy. Provides stronger privacy protection, but may reduce the accuracy of query results; while larger The accuracy of query results is enhanced, but privacy protection may be weaker. Users can choose the appropriate privacy budget according to different scenarios to improve the accuracy of query results as much as possible while ensuring a certain degree of privacy protection.

[0027] Sensitivity Adjustment and privacy budget The choices are as follows: ; in, To query the sensitivity of the task; Transaction data points Query results; Transaction data points Query results; is an adaptive adjustment factor, depending on the characteristics of the data set and query task type ,For example, when the data is unevenly distributed or the query operation is a statistical aggregation, the sensitivity may need to be increased; In traditional differential privacy methods, the sensitivity is usually fixed, and the maximum impact of deleting or adding a data point on the query result is directly calculated. In this innovative solution, we introduce a mechanism to adaptively adjust the sensitivity, taking into account the type of query task and the distribution characteristics of the data, thereby providing a more accurate sensitivity measurement to adapt to different types of data query tasks: ; in, For the original The privacy budget assigned to each query task; is a constant coefficient that controls the basic size of the privacy budget; is the importance weight of the query task, which is used to measure the importance of the query task. , the larger the value, the more important the query task is, because the importance of query tasks varies in actual applications. For example, for key financial data queries, the privacy budget should be higher. This factor can be weighted according to the nature of the query (such as frequent queries, high sensitivity, etc.) to dynamically adjust the privacy budget; Dynamic adjustment factor for query tasks. Dynamic adjustment is performed through error measurement of historical queries (such as changes in query precision) to ensure a balance between privacy budget and query precision. For example, if the previous query results show excessive privacy protection, resulting in inaccurate results, the privacy budget can be appropriately reduced; conversely, the privacy budget can be increased to improve the balance between query precision and privacy protection. is the size of the transaction dataset, which is used to consider the impact of data size on the privacy budget; The sensitivity correction factor for the query type is adjusted according to the type of query task, indicating the difference in privacy requirements for different query tasks (such as aggregation query, count query, etc.). For example, an aggregation query may not require the same strict privacy protection as for each data point, while a frequently updated count query may require more privacy protection. Although It is adjusted dynamically. However, in extreme cases, the privacy budget may be set too low, resulting in unavailability of query results, or too high, leaking too much private information. Therefore, the upper and lower limits of the privacy budget are added to the privacy budget formula for optimization. The optimization is as follows: ; in, For the The privacy budget assigned to each query task; is the minimum value of the privacy budget; is the maximum value of the privacy budget; add upper and lower bound constraints on the privacy budget to ensure that the privacy budget fluctuates within a reasonable range.

[0028] By adding an upper limit to the privacy budget formula and lower limit Constraints to ensure privacy budget It always fluctuates within a reasonable range, thus effectively balancing the relationship between privacy protection and data analysis accuracy. Specifically, when the privacy requirements of the query task are very high, the privacy budget will not be too excessive (to avoid distortion of results caused by overprotection); and when the privacy requirements are low, the privacy budget will not be too loose (to avoid privacy leakage). This optimization enables the system to flexibly and accurately adjust the privacy budget when processing different query tasks, which helps to improve the accuracy of privacy protection and the reliability of query results. At the same time, by preventing the privacy budget from being too high or too low, the potential impact of the system on the query results is reduced, thereby improving the stability and security of the overall data processing process.

[0029] Furthermore, the noise type is selected according to the type of query task and data characteristics, as follows: For simple queries (such as count, minimum, maximum, etc.), choose low-intensity Laplace noise, and the noise parameters can be directly calculated from the sensitivity of the query; For complex queries (such as weighted average, variance, etc.), Gaussian noise is selected to ensure higher privacy protection, and the noise intensity is adjusted considering the distribution characteristics of the data; For high-dimensional data queries, a layered noise addition strategy is adopted to add noise to data of different dimensions separately to optimize computational efficiency and ensure privacy protection.

[0030] The access control unit 4 is used to manage access rights to the stored transaction data according to the user's authority and needs; Ensure that only authorized users can access and decrypt specific data. By setting access control policies, unauthorized users' operations are restricted to protect data security and privacy. After the user enters the requirements, the access control unit 4 verifies the permissions and decides which data can be obtained, decrypted and used for further analysis, while ensuring compliance with data protection and privacy policies; The audit monitoring unit 5 is used to monitor and record the access, operation and change activities of all transaction data in the system in real time. It can also track user behavior, detect abnormal activities, and generate audit reports. Example 2

[0031] The difference between Example 2 of the present invention and Example 1 is that this example introduces the static mechanical properties collection and analysis method used by the trusted data management and control system.

[0032] See also Figure 2 As shown, a trusted data management method is provided, based on the above-mentioned trusted data management system, including the following steps: S1. Collect multi-source user transaction data, identify pre-set sensitive data during the collection process, de-identify the sensitive data, and then encrypt the collected transaction data using a combination of homomorphic encryption and symmetric encryption algorithms; S2. Store the encrypted transaction data and manage access rights to the stored transaction data according to the user's authority and needs; S3. Obtain the transaction data that needs to be analyzed from the stored transaction data according to the user's needs, decrypt the transaction data before obtaining it, and use the differential privacy method to optimize it during the analysis process; S4. Monitor and record all access, operation and change activities of transaction data in the system in real time, track user behavior, detect abnormal activities, and generate audit reports.

[0033] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and descriptions are only preferred examples of the present invention, and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. Trusted data management and control system, characterized by: include: A data input unit (1), the data input unit (1) is used to collect multi-source user transaction data, identify pre-set sensitive data during the collection process, perform de-identification processing on the sensitive data, and transmit the processed transaction data to the encryption protection unit (2); An encryption protection unit (2), the encryption protection unit (2) is used to encrypt the collected transaction data by combining homomorphic encryption and symmetric encryption algorithms, and after encryption is completed, transmit the encrypted transaction data to the storage analysis unit (3); A storage analysis unit (3), the storage analysis unit (3) being used to store the encrypted transaction data, and to obtain transaction data to be analyzed from the stored transaction data according to the requirements input by a user of the access control unit (4), decrypt the transaction data before obtaining the transaction data, and to optimize the analysis process using a differential privacy method; An access control unit (4), the access control unit (4) being used to manage access rights to the stored transaction data according to the rights and requirements of the user; An audit monitoring unit (5), the audit monitoring unit (5) is used to monitor and record in real time the access, operation and change activities of all transaction data in the system, and can also track user behavior, detect abnormal activities, and generate audit reports.

2. The trusted data management and control system according to claim 1, characterized in that: The pre-set sensitive data includes user personal information, user IP address, transaction device ID and transaction amount.

3. The trusted data management and control system according to claim 2, characterized in that: In the encryption protection unit (2), the collected transaction data is encrypted by combining homomorphic encryption and symmetric encryption algorithms. The specific steps are as follows: Use homomorphic encryption for sensitive data: ； in, It is the sensitive data after homomorphic encryption; The original sensitive data; It is a homomorphic encryption operation; Use symmetric encryption for non-sensitive data: ； in, The ciphertext is the symmetrical encryption of non-sensitive data; is non-sensitive data; It is a symmetric encryption algorithm.

4. The trusted data management and control system according to claim 3, characterized in that: The encryption protection unit (2) also includes a key management module (21), which can rotate keys using a hybrid key rotation strategy in a trusted execution environment. At the same time, when a key expires or is threatened with leakage, the key management module (21) triggers a key update process, and the updated key immediately replaces the old key. If the key is leaked or cracked, the key management module (21) can revoke the use of the key.

5. The trusted data management and control system according to claim 4, characterized in that: The specific steps of the key management module (21) rotating the key using the hybrid key rotation strategy in the trusted execution environment are as follows: Based on the audit monitoring unit (5), monitor the usage frequency and real-time risk of the key; Rotation conditions based on frequency of use: ； in, is the key; is the number of times the key is used; is the usage count threshold; The usage time of the key; is the usage duration threshold; Risk-based rotation criteria: ； in, The risk value is calculated based on the real-time risk of transaction data; The risk thresholds to be set; Trigger conditions for the hybrid rotation strategy: ； If any of the conditions is met, key rotation is triggered, and the old key is backed up and a new key is generated, the new key is updated and the old key is destroyed. At the same time, all key rotation operations are recorded and transmitted to the audit monitoring unit (5).

6. The trusted data management and control system according to claim 5, characterized in that: In the storage and analysis unit (3), the specific steps of decrypting the transaction data before acquiring the transaction data are as follows: Use batch operations to aggregate multiple encrypted sensitive data: ； in, is the total sensitive data that has been homomorphically encrypted; For each sensitive data Homomorphic encryption result of ; The amount of sensitive data; ; Represents a batch homomorphic addition operation; Homomorphically decrypt the encrypted total sensitive data: ； in, The total sensitive data for homomorphic encryption Decryption is performed, and the decrypted result is the total transaction amount in plain text; is the total sensitive data after decryption; Symmetrically decrypt sensitive data: ； in, For non-sensitive data after symmetrical encryption Decrypt and obtain the plaintext data; It is the ciphertext data after symmetrical encryption; The decrypted sensitive data.

7. The trusted data management and control system according to claim 6, characterized in that: In the storage analysis unit (3), a differential privacy method is used for optimization during the analysis process, and the optimization is as follows: First, determine the query operation entered by the user and adjust the query sensitivity according to the type of transaction data query task. ; Then identify the importance of the query task, based on the importance of the query task and the sensitivity of the query and dynamically allocate privacy budgets based on expected privacy needs ; Then, the noise type is selected according to the type of query task and the characteristics of transaction data, and the noise is added to the query process, so as to perform the query and output the result after adding noise.

8. The trusted data management and control system according to claim 7, characterized in that: The sensitivity Adjustment and privacy budget The choices are as follows: ； in, To query the sensitivity of the task; Transaction data points Query results; Transaction data points Query results; is the adaptive adjustment factor; ； in, For the original The privacy budget assigned to each query task; is a constant coefficient; is the importance weight of the query task; is the dynamic adjustment factor of the query task; is the size of the transaction dataset; Is the sensitivity correction factor for the query type.

9. The trusted data management and control system according to claim 8, characterized in that: The upper and lower limits of the privacy budget are added to the privacy budget formula for optimization. The optimization is as follows: ； in, For the The privacy budget assigned to each query task; is the minimum value of the privacy budget; is the maximum value of the privacy budget.

10. A trusted data management method, based on the trusted data management system according to any one of claims 1 to 9, characterized in that: The steps include: S1. Collect multi-source user transaction data, identify pre-set sensitive data during the collection process, de-identify the sensitive data, and then encrypt the collected transaction data using a combination of homomorphic encryption and symmetric encryption algorithms; S2. Store the encrypted transaction data and manage access rights to the stored transaction data according to the user's authority and needs; S3. Obtain the transaction data that needs to be analyzed from the stored transaction data according to the user's needs, decrypt the transaction data before obtaining it, and use the differential privacy method to optimize it during the analysis process; S4. Monitor and record all access, operation and change activities of transaction data in the system in real time, track user behavior, detect abnormal activities, and generate audit reports.

Images