Private data analysis method and system based on collaborative learning and dynamic encryption

Abstract

The invention provides a privacy data analysis method and system based on collaborative learning and dynamic encryption, and relates to the technical field of privacy, and the method comprises the steps: receiving original data of a plurality of participants, the original data of each participant being in an initial encryption state; performing secondary encryption on the original data of each participant by using the dynamic encryption key to obtain secondary encryption data; inputting the secondary encrypted data into a preset collaborative learning model, and decrypting the collaborative learning model by using a security computing protocol under homomorphic encryption in combination with a dynamic encryption key of each participant to obtain a privacy protection analysis model which can be used by each participant; when data analysis is carried out, the privacy protection analysis model is used for processing input data to be analyzed, and correctness of the analysis process is ensured through a verifiable computing technology; and based on a predefined access control strategy, only returning an analysis result that the participant has access to the participant to the corresponding participant.

Description

Technical Field

[0001] The present invention relates to privacy technology, and in particular to a privacy data analysis method and system based on collaborative learning and dynamic encryption. Background Art

[0002] With the advent of the big data era, data analysis is increasingly used in various fields. However, data privacy protection issues have also become prominent and have become an important issue that needs to be addressed. Traditional data analysis methods often require centralized collection and processing of large amounts of raw data, which not only increases the risk of data leakage, but also faces challenges in terms of laws and regulations.

[0003] Most existing privacy protection schemes use static security strategies, which are difficult to adapt to different application scenarios and dynamically changing privacy needs. In practical applications, the sensitivity of data may change over time, environment and other factors, and a fixed privacy protection level may lead to over-protection or under-protection.

[0004] Therefore, a new method is urgently needed that can protect data privacy while maintaining high data analysis efficiency and accuracy. This method should be able to adapt to large-scale, heterogeneous data environments and dynamically adjust the privacy protection level according to actual needs to balance the relationship between privacy protection and data utility. Summary of the invention

[0005] The embodiments of the present invention provide a privacy data analysis method and system based on collaborative learning and dynamic encryption, which can solve the problems in the prior art.

[0006] According to a first aspect of the embodiments of the present invention, Provides a privacy data analysis method based on collaborative learning and dynamic encryption, including: Receive original data from multiple participants, where the original data of each participant is in an initial encryption state, and the initial encryption state is implemented using a homomorphic encryption algorithm; assign a unique participant identifier to each participant, and the participant identifier includes a timestamp and a random hash value; generate a corresponding dynamic encryption key based on the participant identifier, and the dynamic encryption key is generated using quantum key distribution technology; use the dynamic encryption key to perform secondary encryption on the original data of each participant to obtain secondary encrypted data, and the secondary encryption adopts a layered encryption strategy, and uses encryption algorithms of different strengths for different sensitivity levels of data; Input the secondary encrypted data into a preset collaborative learning model, wherein the collaborative learning model includes multiple sub-models, each sub-model corresponds to a participant, and the sub-model adopts a federated learning architecture; utilize a secure computing protocol under homomorphic encryption and a dynamic encryption key of each participant to decrypt the collaborative learning model, thereby obtaining a privacy protection analysis model that can be used by each participant; When performing data analysis, the privacy protection analysis model is used to process the input data to be analyzed, and the correctness of the analysis process is ensured through verifiable computing technology; based on the predefined access control policy, only the analysis results that the corresponding participants are entitled to access are returned, and the analysis results are k-anonymized to further enhance privacy protection; the model update and retraining process is regularly performed to adapt to the dynamic changes in data distribution.

[0007] In an optional embodiment, Generating a corresponding dynamic encryption key based on the participant identifier, wherein the dynamic encryption key is generated using quantum key distribution technology, includes: An entangled photon pair is generated by a periodically poled lithium niobate crystal through a spontaneous parametric down-conversion process, wherein the temperature of the periodically poled lithium niobate crystal is controlled at 163.5±0.1°C, and the wavelength of the entangled photon pair is 1550nm; one photon in the entangled photon pair is sent to a sender, and the other photon is sent to a receiver, wherein the photon transmission adopts a photonic crystal fiber with an optical loss rate not higher than 0.2dB / km; a quantum repeater is arranged every 50km on the transmission path of the photonic crystal fiber, wherein the quantum repeater adopts a rare earth-doped YSO crystal as a quantum storage medium, and the storage time is not less than 1ms; The sender and the receiver respectively use a superconducting nanowire single-photon detector to measure the quantum state of the received photons, wherein the quantum efficiency of the superconducting nanowire single-photon detector at a wavelength of 1550nm is not less than 93%, and the dark count rate is not higher than 1Hz; Based on the quantum state measurement result, a quantum key is generated through a post-processing step, which includes basis comparison, parameter estimation, information coordination and privacy amplification, wherein information coordination adopts a rate-adaptive low-density parity-check code, and privacy amplification adopts a privacy matrix with dynamically adjustable size.

[0008] In an optional embodiment, The original data of each participant is re-encrypted using the dynamic encryption key to obtain re-encrypted data. The re-encryption adopts a layered encryption strategy, and encryption algorithms of different strengths are adopted for different sensitivity levels of data, including: A data sensitivity assessment model based on deep learning is used to perform sensitivity assessment on the encrypted data, wherein the data sensitivity assessment model based on deep learning includes 12 base layers, 2 fully connected layers, 1 Dropout layer and 1 output layer, wherein the dropout rate of the Dropout layer is 0.1, and the output layer uses a Sigmoid activation function; Based on the quantum key, three sub-keys are generated using a key derivation function, which are used for data encryption at three sensitivity levels: high, medium and low; For high-sensitivity data with a sensitivity score greater than 0.8, the Kyber1024 algorithm is used for key encapsulation, and AES-256-GCM is used for authenticated encryption; for medium-sensitivity data with a sensitivity score between 0.4 and 0.8, the NTRU-HRSS-701 parameter set is used for key generation and encapsulation, and ChaCha20-Poly1305 is used for authenticated encryption; for low-sensitivity data with a sensitivity score less than 0.4, the derived subkey is directly used, and XChaCha20-Poly1305 is used for authenticated encryption.

[0009] In an optional embodiment, Before inputting the secondary encrypted data into the preset collaborative learning model, the method further includes training the collaborative learning model: During the training process of the collaborative learning model, each sub-model only processes the secondary encrypted data of the participant and generates an intermediate result, which includes the model parameter gradient and the local loss function value; a secure multi-party computing protocol is used to collect the intermediate results generated by each sub-model, and the intermediate results are aggregated to form a global model, and the aggregation process adopts differential privacy technology to prevent the reverse deduction of the original data; the global model is used to perform analysis and processing on the newly input data to be analyzed under privacy protection, and the analysis and processing includes feature extraction, classification, regression and clustering; during the training process of the collaborative learning model, the dynamic encryption key of each participant is periodically updated based on the Bloom filter and the time decay function, and the original data of each participant is re-encrypted using the updated dynamic encryption key; Monitoring the training process of the collaborative learning model, and triggering a model verification procedure when it is detected that the model performance reaches a preset threshold or a potential model poisoning attack is detected; The model verification procedure includes: using zero-knowledge proof technology to verify the integrity of the model, using adversarial sample testing to evaluate the robustness of the model, and ensuring the fairness of the model through multi-party consistency checks; dynamically adjusting the collaborative learning strategy based on the model verification results, including updating the weights of the participants, adjusting the learning rate, and modifying the model structure; when the model verification passes and the performance meets the requirements, the training process is terminated and the final collaborative learning model is output.

[0010] In an optional embodiment, During the training process of the collaborative learning model, each sub-model processes only the secondary encrypted data of the participant and generates intermediate results, which include model parameter gradients and local loss function values; the intermediate results generated by each sub-model are collected using a secure multi-party computing protocol, and the intermediate results are aggregated to form a global model, and the aggregation process uses differential privacy technology to prevent reverse deduction of original data; the global model is used to perform analysis and processing on the newly input data to be analyzed under privacy protection, and the analysis and processing includes feature extraction, classification, regression and clustering, including: The original data of the participants is encrypted using the CKKS approximate homomorphic encryption scheme, wherein the polynomial modulus of the CKKS approximate homomorphic encryption scheme is 2^15, the plaintext modulus is 2^40, the ciphertext modulus is approximately 2^440, and the security level is greater than 128 bits; in the local sub-model of each participant, the encrypted data is batch encoded, multiple real numbers are packaged into a polynomial, and the nonlinear activation function is approximated using the Taylor expansion polynomial; forward propagation and back propagation calculations are performed on the encrypted data to generate encrypted model parameter gradients and encrypted local loss function values; Generate a public key and a private key using the (T,n) threshold Paillier encryption scheme, where T is the reconstruction threshold and n is the number of participants. Use the Shamir secret sharing scheme to split the private key into n parts and distribute them to each participant. Each participant uses the public key to encrypt its model parameter gradient and local loss function value, and adds calibrated Gaussian noise before encryption. The noise variance σ² of the calibrated Gaussian noise is calculated based on the preset privacy budget ε and sensitivity Δf. One participant is selected as an aggregator, and the other participants send the encrypted intermediate results to the aggregator, and the aggregator calculates the sum of the encrypted gradient and the encrypted loss; at least n participants cooperate to partially decrypt the encrypted total gradient and total loss using their respective private key shares, and the aggregator collects the partial decryption results and completes the final decryption to obtain the global gradient and global loss; The global model parameters are updated using the global gradient, and the global model is used to perform analysis and processing of newly input data to be analyzed under privacy protection; in the analysis and processing, the SimHash algorithm is used to extract features and map high-dimensional feature vectors to low-dimensional binary space; an exponential mechanism is used to select the best splitting features to construct a privacy-preserving decision tree ensemble; inner product function encryption is used to realize privacy-preserving linear regression; and Laplace noise is introduced in each iteration of the K-means clustering algorithm to protect the coordinates of the cluster center.

[0011] In an optional embodiment, During the training process of the collaborative learning model, the dynamic encryption key of each participant is periodically updated based on the Bloom filter and the time decay function, and the original data of each participant is re-encrypted using the updated dynamic encryption key, including: Set a Bloom filter whose size m and number of hash functions k are calculated based on the expected number of elements N and the false positive probability p; use the exponential decay function w(t) = exp(-λt) to calculate the key update weight, where λ is the decay rate and t is the key usage time; For each encryption operation, its characteristic hash is calculated and inserted into the Bloom filter, and the filling rate of the Bloom filter is periodically checked; when the filling rate exceeds a preset filling threshold, a key update is triggered, a new key is generated, and the exponential decay function is used to calculate the mixing weight, the new and old keys are mixed to obtain the final key, and the updated dynamic encryption key is used to re-encrypt the original data of each participant.

[0012] According to a second aspect of the embodiments of the present invention, Provides a privacy data analysis system based on collaborative learning and dynamic encryption, including: The first unit is used to receive original data of multiple participants, where the original data of each participant is in an initial encryption state, and the initial encryption state is implemented by a homomorphic encryption algorithm; a unique participant identifier is assigned to each participant, and the participant identifier includes a timestamp and a random hash value; a corresponding dynamic encryption key is generated based on the participant identifier, and the dynamic encryption key is generated by a quantum key distribution technology; the original data of each participant is secondary encrypted by using the dynamic encryption key to obtain secondary encrypted data, and the secondary encryption adopts a layered encryption strategy, and encryption algorithms of different strengths are used for different sensitivity levels of data; The second unit is used to input the secondary encrypted data into a preset collaborative learning model, wherein the collaborative learning model includes multiple sub-models, each sub-model corresponds to a participant, and the sub-model adopts a federated learning architecture; the collaborative learning model is decrypted by using a secure computing protocol under homomorphic encryption and a dynamic encryption key of each participant to obtain a privacy protection analysis model that can be used by each participant; The third unit is used to process the input data to be analyzed by using the privacy protection analysis model during data analysis, and ensure the correctness of the analysis process through verifiable computing technology; based on the predefined access control policy, only the analysis results that the corresponding participants are entitled to access are returned, and the analysis results are k-anonymized to further enhance privacy protection; and the model update and retraining process are regularly executed to adapt to the dynamic changes in data distribution.

[0013] According to a third aspect of the embodiments of the present invention, An electronic device is provided, comprising: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0014] According to a fourth aspect of the embodiments of the present invention, A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the aforementioned method is implemented.

[0015] The quantum key generated by this application has a high degree of security and randomness and can be used in the subsequent dynamic encryption process. This dynamic encryption key generation method based on quantum key distribution technology combines the basic principles of quantum physics and advanced information processing technology, providing a solid security foundation for privacy data analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flowchart of a privacy data analysis method based on collaborative learning and dynamic encryption according to an embodiment of the present invention; Figure 2 This is a structural diagram of a privacy data analysis system based on collaborative learning and dynamic encryption according to an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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.

[0018] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0019] Figure 1 FIG. 1 is a flow chart of a privacy data analysis method based on collaborative learning and dynamic encryption according to an embodiment of the present invention. Figure 1 As shown, the method includes: S101. Receive original data from multiple participants, where the original data of each participant is in an initial encryption state, and the initial encryption state is implemented using a homomorphic encryption algorithm; assign a unique participant identifier to each participant, where the participant identifier includes a timestamp and a random hash value; generate a corresponding dynamic encryption key based on the participant identifier, and the dynamic encryption key is generated using quantum key distribution technology; use the dynamic encryption key to perform secondary encryption on the original data of each participant to obtain secondary encrypted data, and the secondary encryption adopts a layered encryption strategy, and uses encryption algorithms of different strengths for different sensitivity levels of data; S102. Input the secondary encrypted data into a preset collaborative learning model, wherein the collaborative learning model includes a plurality of sub-models, each sub-model corresponds to a participant, and the sub-model adopts a federated learning architecture; decrypt the collaborative learning model using a secure computing protocol under homomorphic encryption and a dynamic encryption key of each participant, thereby obtaining a privacy protection analysis model that can be used by each participant; S103. When performing data analysis, the privacy-preserving analysis model is used to process the input data to be analyzed, and the correctness of the analysis process is ensured through verifiable computing technology. Based on the predefined access control policy, only the analysis results that the corresponding participants are entitled to access are returned, and the analysis results are k-anonymized to further enhance privacy protection. The model update and retraining process is performed regularly to adapt to the dynamic changes in data distribution.

[0020] In an optional embodiment, Generating a corresponding dynamic encryption key based on the participant identifier, wherein the dynamic encryption key is generated using quantum key distribution technology, includes: An entangled photon pair is generated by a periodically poled lithium niobate crystal through a spontaneous parametric down-conversion process, wherein the temperature of the periodically poled lithium niobate crystal is controlled at 163.5±0.1°C, and the wavelength of the entangled photon pair is 1550nm; one photon in the entangled photon pair is sent to a sender, and the other photon is sent to a receiver, wherein the photon transmission adopts a photonic crystal fiber with an optical loss rate not higher than 0.2dB / km; a quantum repeater is arranged every 50km on the transmission path of the photonic crystal fiber, wherein the quantum repeater adopts a rare earth-doped YSO crystal as a quantum storage medium, and the storage time is not less than 1ms; The sender and the receiver respectively use a superconducting nanowire single-photon detector to measure the quantum state of the received photons, wherein the quantum efficiency of the superconducting nanowire single-photon detector at a wavelength of 1550nm is not less than 93%, and the dark count rate is not higher than 1Hz; Based on the quantum state measurement result, a quantum key is generated through a post-processing step, which includes basis comparison, parameter estimation, information coordination and privacy amplification, wherein information coordination adopts a rate-adaptive low-density parity-check code, and privacy amplification adopts a privacy matrix with dynamically adjustable size.

[0021] Exemplarily, first, a periodically poled lithium niobate crystal is prepared as a source for generating entangled photon pairs. The crystal temperature is precisely controlled at 163.5±0.1°C, a temperature range that is critical for maintaining the nonlinear optical properties of the crystal. Under this condition, entangled photon pairs with a wavelength of 1550nm are generated through a spontaneous parametric down-conversion process. The 1550nm wavelength is chosen because it has low loss in optical fiber transmission and is suitable for long-distance quantum communication.

[0022] Next, the generated entangled photon pair is separated, with one photon sent to the sender and the other to the receiver. The photons are transmitted using specially designed photonic crystal fibers, whose optical loss rate does not exceed 0.2dB / km. This low-loss characteristic can significantly improve the transmission distance and quality of quantum signals. For example, over a transmission distance of 100km, the signal strength is only attenuated by about 20%, which is much better than traditional optical fibers.

[0023] Considering the attenuation of quantum signals in long-distance transmission, a quantum repeater is set up every 50km on the transmission path of photonic crystal fiber. These repeaters use rare earth-doped YSO crystals as quantum storage media, which can achieve a storage time of no less than 1ms. This means that quantum information can be temporarily stored in the repeater for at least 1 millisecond, providing a sufficient time window for subsequent quantum state operations and transmission.

[0024] When the photons reach the sender and receiver, both parties use superconducting nanowire single-photon detectors to measure the quantum state. The selected detector has a quantum efficiency of at least 93% at a wavelength of 1550nm, and the dark count rate is less than 1Hz. This high-efficiency, low-noise detector can accurately capture weak quantum signals and ensure the reliability of the measurement results. For example, within 1 second, the detector can effectively identify more than 93,000 photons, and no more than 1 false detection signal is generated.

[0025] Based on the results of quantum state measurement, the sender and receiver perform a series of post-processing steps to generate the final quantum key. The first is basis comparison, where both parties publicly communicate their respective measurement basis choices and retain measurement results that are consistent with the basis. Then parameter estimation is performed to evaluate the error rate of the quantum channel and the possible amount of information leakage.

[0026] Next is the information coordination stage, where rate-adaptive low-density parity-check codes are used to correct errors in the measurement results. This coding method can dynamically adjust the coding rate according to the actual channel conditions, minimizing information leakage while ensuring error correction. For example, when the channel error rate is 5%, a coding rate of 0.8 may be used; when the error rate rises to 10%, the coding rate may be reduced to 0.6 to provide stronger error correction capabilities.

[0027] The final privacy amplification step uses a privacy matrix with dynamically adjustable size. The dimensions of this matrix can change dynamically according to the current security requirements and computing resources. For example, under high security requirements, a large matrix of 10000×5000 may be used; while under resource constraints, it may be reduced to a size of 1000×500. By adjusting the matrix size, a balance can be achieved between security and efficiency.

[0028] After this series of steps, the quantum key finally generated has a high degree of security and randomness, and can be used in the subsequent dynamic encryption process. This dynamic encryption key generation method based on quantum key distribution technology combines the basic principles of quantum physics and advanced information processing technology, providing a solid security foundation for privacy data analysis.

[0029] In an optional embodiment, The original data of each participant is re-encrypted using the dynamic encryption key to obtain re-encrypted data. The re-encryption adopts a layered encryption strategy, and encryption algorithms of different strengths are adopted for different sensitivity levels of data, including: A data sensitivity assessment model based on deep learning is used to perform sensitivity assessment on the encrypted data, wherein the data sensitivity assessment model based on deep learning includes 12 base layers, 2 fully connected layers, 1 Dropout layer and 1 output layer, wherein the dropout rate of the Dropout layer is 0.1, and the output layer uses a Sigmoid activation function; Based on the quantum key, three sub-keys are generated using a key derivation function, which are used for data encryption at three sensitivity levels: high, medium and low; For high-sensitivity data with a sensitivity score greater than 0.8, the Kyber1024 algorithm is used for key encapsulation, and AES-256-GCM is used for authenticated encryption; for medium-sensitivity data with a sensitivity score between 0.4 and 0.8, the NTRU-HRSS-701 parameter set is used for key generation and encapsulation, and ChaCha20-Poly1305 is used for authenticated encryption; for low-sensitivity data with a sensitivity score less than 0.4, the derived subkey is directly used, and XChaCha20-Poly1305 is used for authenticated encryption.

[0030] Exemplarily, first, the sensitivity of the encrypted data is assessed. This step uses a data sensitivity assessment model based on deep learning, and the model architecture includes 12 base layers, 2 fully connected layers, 1 Dropout layer, and 1 output layer. The base layer is responsible for feature extraction and can be a convolutional layer or a recurrent layer, depending on the data type. The fully connected layer is used to map the extracted features to a sensitivity score. The dropout rate of the Dropout layer is set to 0.1 to help prevent overfitting. The output layer uses a Sigmoid activation function to map the final result to between 0 and 1, representing the sensitivity score of the data.

[0031] For example, for a piece of data containing name, age, address, and medical records, the model might give a sensitivity score of 0.95, indicating that this is highly sensitive information, while for data containing only age and city, the model might give a score of 0.3, indicating a lower sensitivity.

[0032] Next, based on the previously generated quantum key, a key derivation function is used to generate three subkeys. This step ensures that even if one subkey is cracked, the other subkeys remain secure. For example, HKDF (HMAC-basedKey Derivation Function) can be used as a key derivation function to derive three 256-bit subkeys from a 256-bit quantum key, which are used to encrypt data at three sensitivity levels: high, medium, and low.

[0033] Subsequently, the corresponding encryption algorithm is selected based on the sensitivity score of the data. For highly sensitive data with a sensitivity score greater than 0.8, the Kyber1024 algorithm is used for key encapsulation. Kyber1024 is a post-quantum cryptography algorithm that is resistant to attacks by quantum computers. The encapsulated key is used in the AES-256-GCM algorithm for authenticated encryption. This combination provides extremely high security and is suitable for protecting the most sensitive data.

[0034] For medium-sensitivity data with a sensitivity score between 0.4 and 0.8, the NTRU-HRSS-701 parameter set is used for key generation and encapsulation. NTRU is also a post-quantum cryptography algorithm that provides strong security. The encapsulated keys are used for authenticated encryption with the ChaCha20-Poly1305 algorithm. This combination strikes a good balance between security and performance and is suitable for processing medium-sensitivity data.

[0035] For low-sensitivity data with a sensitivity score less than 0.4, the derived subkey is directly used to perform authenticated encryption using XChaCha20-Poly1305. XChaCha20 is an extended version of ChaCha20 that uses a longer nonce for enhanced security. This approach provides adequate protection for low-sensitivity data while maintaining high encryption efficiency.

[0036] In practical applications, for example, for a data entry containing name, ID number, and bank account information, the sensitivity score is 0.92. The system will use Kyber1024 to encapsulate a temporary key and then encrypt the data using AES-256-GCM. The encryption process not only ensures the confidentiality of the data, but also provides integrity protection through GCM mode.

[0037] For a data entry containing age, occupation, and city of work, the sensitivity score is 0.6. The system generates and wraps the key using NTRU-HRSS-701, and then encrypts the data using ChaCha20-Poly1305. This method provides faster encryption speed than encryption methods for highly sensitive data while ensuring security.

[0038] For data entries that only contain public information such as company name and year of establishment, the sensitivity score is 0.2. The system directly uses the derived subkey and XChaCha20-Poly1305 for encryption. This method provides fast encryption speed and is suitable for processing large amounts of low-sensitivity data.

[0039] Through this layered encryption strategy, the system can flexibly select encryption strength according to the sensitivity of the data, optimizing the use of computing resources while ensuring security. This approach not only improves overall encryption efficiency, but also ensures that the most sensitive data receives the strongest protection.

[0040] In an optional embodiment, Before inputting the secondary encrypted data into the preset collaborative learning model, the method further includes training the collaborative learning model: During the training process of the collaborative learning model, each sub-model only processes the secondary encrypted data of the participant and generates an intermediate result, which includes the model parameter gradient and the local loss function value; a secure multi-party computing protocol is used to collect the intermediate results generated by each sub-model, and the intermediate results are aggregated to form a global model, and the aggregation process adopts differential privacy technology to prevent the reverse deduction of the original data; the global model is used to perform analysis and processing on the newly input data to be analyzed under privacy protection, and the analysis and processing includes feature extraction, classification, regression and clustering; during the training process of the collaborative learning model, the dynamic encryption key of each participant is periodically updated based on the Bloom filter and the time decay function, and the original data of each participant is re-encrypted using the updated dynamic encryption key; Monitoring the training process of the collaborative learning model, and triggering a model verification procedure when it is detected that the model performance reaches a preset threshold or a potential model poisoning attack is detected; The model verification procedure includes: using zero-knowledge proof technology to verify the integrity of the model, using adversarial sample testing to evaluate the robustness of the model, and ensuring the fairness of the model through multi-party consistency checks; dynamically adjusting the collaborative learning strategy based on the model verification results, including updating the weights of the participants, adjusting the learning rate, and modifying the model structure; when the model verification passes and the performance meets the requirements, the training process is terminated and the final collaborative learning model is output.

[0041] For example, before the secondary encrypted data is input into the preset collaborative learning model, the model needs to be trained. This training process involves multiple key steps to ensure the performance, security and privacy protection of the model.

[0042] First, during the training of the collaborative learning model, each participant's sub-model only processes its own secondary encrypted data. For example, for a medical data analysis task, the sub-model of hospital A may process 1,000 encrypted patient records, while the sub-model of hospital B processes 800. Each sub-model generates intermediate results after processing, including model parameter gradients and local loss function values. These intermediate results do not directly expose the original data, but reflect the progress of model learning.

[0043] Next, a secure multi-party computation protocol is used to collect the intermediate results generated by each sub-model. This step ensures the security of the intermediate results during transmission and aggregation. For example, homomorphic encryption technology can be used to enable each party to perform calculations in an encrypted state without decrypting the intermediate results.

[0044] Then, the collected intermediate results are aggregated to form a global model. During the aggregation process, differential privacy techniques are applied to prevent reverse deduction of the original data. Specifically, calibrated noise can be added to the gradient before aggregation. For example, for a gradient with a sensitivity of 1, Gaussian noise with a mean of 0 and a standard deviation of 0.1 may be added, which protects privacy without significantly affecting model performance.

[0045] The global model formed can be used to analyze and process the newly input data to be analyzed under privacy protection. This includes tasks such as feature extraction, classification, regression, and clustering. For example, in a medical diagnosis scenario, the model may extract key features from patient data, classify diseases, or predict treatment effects.

[0046] During the training process of the collaborative learning model, the dynamic encryption keys of each participant need to be updated regularly. This step is implemented based on the Bloom filter and the time decay function. The Bloom filter is used to quickly check whether the key needs to be updated, while the time decay function determines the frequency of the update. For example, the key can be set to be updated every 24 hours, or when a location in the Bloom filter is accessed more than 1,000 times, the update is triggered. The updated dynamic encryption key is used to re-encrypt the original data of each participant, further enhancing data security.

[0047] During the entire training process, the performance of the collaborative learning model needs to be continuously monitored. When the model performance reaches a preset threshold (for example, the accuracy exceeds 95%) or a potential model poisoning attack is discovered (such as a sudden drop in model accuracy of more than 10%), the model verification procedure is triggered.

[0048] The model verification process includes multiple steps. First, the integrity of the model is verified using zero-knowledge proof technology. This ensures that the model has not been tampered with and that all parties have followed the protocol. For example, the zk-SNARK protocol can be used to allow each party to prove that they have correctly performed the model update without revealing the specific content of the update.

[0049] Second, the robustness of the model is evaluated using adversarial sample testing. This involves generating carefully designed inputs that try to fool the model. For example, in an image classification task, small perturbations may be added to normal images to check whether the model can still classify them correctly. If the model maintains correct predictions on more than 95% of adversarial examples, it can be considered to have good robustness.

[0050] Additionally, ensure model fairness through multi-party consistency checks. This may include checking that the model performs consistently across different demographic groups. For example, in a loan approval model, ensure that different applicants are treated fairly, with a difference of no more than 5%.

[0051] Dynamically adjust the collaborative learning strategy based on the results of model validation. This may include updating participant weights (such as reducing the weight of underperforming participants), adjusting the learning rate (such as reducing the learning rate from 0.01 to 0.001 when model performance stagnates), or modifying the model structure (such as adding or reducing convolutional layers in a convolutional neural network).

[0052] When the model passes the validation and the performance meets the requirements, the training process is terminated and the final collaborative learning model is output. For example, when the accuracy of the model on the validation set exceeds 98% and passes all security and fairness checks, the training can be considered complete.

[0053] This complete training process not only ensures the performance of the collaborative learning model, but also maintains data privacy and model security throughout the process. Through periodic key updates, continuous monitoring and verification, and dynamic policy adjustments, this method can cope with various potential security threats and performance challenges, and ultimately produce a high-quality, trustworthy collaborative learning model.

[0054] In an optional embodiment, During the training process of the collaborative learning model, each sub-model processes only the secondary encrypted data of the participant and generates intermediate results, which include model parameter gradients and local loss function values; the intermediate results generated by each sub-model are collected using a secure multi-party computing protocol, and the intermediate results are aggregated to form a global model, and the aggregation process uses differential privacy technology to prevent reverse deduction of original data; the global model is used to perform analysis and processing on the newly input data to be analyzed under privacy protection, and the analysis and processing includes feature extraction, classification, regression and clustering, including: The original data of the participants is encrypted using the CKKS approximate homomorphic encryption scheme, wherein the polynomial modulus of the CKKS approximate homomorphic encryption scheme is 2^15, the plaintext modulus is 2^40, the ciphertext modulus is approximately 2^440, and the security level is greater than 128 bits; in the local sub-model of each participant, the encrypted data is batch encoded, multiple real numbers are packaged into a polynomial, and the nonlinear activation function is approximated using the Taylor expansion polynomial; forward propagation and back propagation calculations are performed on the encrypted data to generate encrypted model parameter gradients and encrypted local loss function values; Generate a public key and a private key using the (T,n) threshold Paillier encryption scheme, where T is the reconstruction threshold and n is the number of participants. Use the Shamir secret sharing scheme to split the private key into n parts and distribute them to each participant. Each participant uses the public key to encrypt its model parameter gradient and local loss function value, and adds calibrated Gaussian noise before encryption. The noise variance σ² of the calibrated Gaussian noise is calculated based on the preset privacy budget ε and sensitivity Δf. One participant is selected as an aggregator, and the other participants send the encrypted intermediate results to the aggregator, and the aggregator calculates the sum of the encrypted gradient and the encrypted loss; at least n participants cooperate to partially decrypt the encrypted total gradient and total loss using their respective private key shares, and the aggregator collects the partial decryption results and completes the final decryption to obtain the global gradient and global loss; The global model parameters are updated using the global gradient, and the global model is used to perform analysis and processing of newly input data to be analyzed under privacy protection; in the analysis and processing, the SimHash algorithm is used to extract features and map high-dimensional feature vectors to low-dimensional binary space; an exponential mechanism is used to select the best splitting features to construct a privacy-preserving decision tree ensemble; inner product function encryption is used to realize privacy-preserving linear regression; and Laplace noise is introduced in each iteration of the K-means clustering algorithm to protect the coordinates of the cluster center.

[0055] For example, during the training of the collaborative learning model, each participant's sub-model only processes its own secondary encrypted data and generates intermediate results. This process involves multiple complex technical steps to ensure data privacy and model security.

[0056] First, the original data of the participants is encrypted using the CKKS approximate homomorphic encryption scheme. The polynomial modulus of the CKKS scheme is set to 2^15, the plaintext modulus is 2^40, and the ciphertext modulus is approximately 2^440. The selection of these parameters ensures a security level greater than 128 bits. For example, for a medical data set containing a patient's age, blood pressure, and blood sugar level, each value is converted into an encrypted polynomial.

[0057] In each participant's local sub-model, the encrypted data is batch encoded. This step packs multiple real numbers into a polynomial, greatly improving computational efficiency. For example, the age data of 10 patients may be packed into a polynomial. For nonlinear activation functions such as ReLU or Sigmoid, Taylor expansion polynomial approximation is used. This allows approximate nonlinear operations in the encrypted domain.

[0058] Next, forward propagation and back propagation calculations are performed on the encrypted data. In forward propagation, the encrypted input data passes through each layer of the model to generate a prediction result. In back propagation, the gradient of the loss function to the parameters of each layer is calculated. All these operations are performed in an encrypted state, and finally the encrypted model parameter gradients and encrypted local loss function values ​​are generated.

[0059] In order to securely aggregate these encrypted intermediate results, the (T,n) threshold Paillier encryption scheme is introduced. Here, T is the reconstruction threshold and n is the number of participants. For example, in the case of 5 participants, T=3 may be set, which means that at least 3 participants are required to cooperate to decrypt the data. The private key is split into n parts using the Shamir secret sharing scheme and distributed to each participant. Each participant gets a private key fragment and cannot decrypt the data alone.

[0060] Each participant uses a public key to encrypt its model parameter gradients and local loss function values. Before encryption, calibrated Gaussian noise is added to achieve differential privacy. The variance of the noise is calculated based on the preset privacy budget and sensitivity. For example, if the privacy budget ε is set to 0.1 and the sensitivity is 1, Gaussian noise with a standard deviation of about 10 may be added.

[0061] In the aggregation phase, one party is selected as the aggregator. The other parties send the encrypted intermediate results to the aggregator. The aggregator calculates the sum of the encrypted gradient and the encrypted loss, but due to the nature of homomorphic encryption, the aggregator cannot see the actual value.

[0062] At least T participants (3 in the above example) cooperate to partially decrypt the encrypted total gradient and total loss using their respective private key shares. The aggregator collects these partial decryption results and completes the final decryption to obtain the global gradient and global loss. This process ensures that the aggregated result can only be obtained when enough participants cooperate.

[0063] The decrypted global gradient is used to update the global model parameters. This updated global model is used to analyze and process the newly input data to be analyzed under privacy protection.

[0064] In the analysis and processing stage, the SimHash algorithm is first used for feature extraction. SimHash can map high-dimensional feature vectors to low-dimensional binary space, greatly reducing storage and computing requirements. For example, a 1000-dimensional patient feature vector may be compressed into a 64-bit binary string.

[0065] For classification tasks, an exponential mechanism is used to select the best split features to build a privacy-preserving decision tree ensemble. The exponential mechanism randomly selects split features based on the information gain and privacy budget of the features, which ensures both model performance and data privacy.

[0066] In the regression task, inner product function encryption is used to achieve privacy-preserving linear regression. This allows the inner product of the feature vector and the weight vector to be calculated in an encrypted state, thereby making predictions without exposing the original data.

[0067] For clustering tasks, Laplace noise is introduced in each iteration of the K-means algorithm to protect the coordinates of the cluster centers. For example, for two-dimensional data, Laplace noise with a mean of 0 and a scale parameter of 1 may be added to the x and y coordinates of each cluster center, respectively.

[0068] This whole process forms a closed loop: from encryption of raw data, to training of local models, to secure aggregation, and finally to updating and applying the global model. Each step is built on the previous step and provides necessary input for the next step. By comprehensively using technologies such as homomorphic encryption, secret sharing, and differential privacy, this method achieves efficient collaborative learning and data analysis while protecting data privacy.

[0069] In an optional embodiment, During the training process of the collaborative learning model, the dynamic encryption key of each participant is periodically updated based on the Bloom filter and the time decay function, and the original data of each participant is re-encrypted using the updated dynamic encryption key, including: Set a Bloom filter whose size m and number of hash functions k are calculated based on the expected number of elements N and the false positive probability p; use the exponential decay function w(t) = exp(-λt) to calculate the key update weight, where λ is the decay rate and t is the key usage time; For each encryption operation, its characteristic hash is calculated and inserted into the Bloom filter, and the filling rate of the Bloom filter is periodically checked; when the filling rate exceeds a preset filling threshold, a key update is triggered, a new key is generated, and the exponential decay function is used to calculate the mixing weight, the new and old keys are mixed to obtain the final key, and the updated dynamic encryption key is used to re-encrypt the original data of each participant.

[0070] For example, in the training process of the collaborative learning model, in order to ensure data security and model performance, it is necessary to periodically update the dynamic encryption keys of each participant. This process is implemented based on Bloom filters and time decay functions, and specifically involves multiple technical steps to form a closed-loop key management mechanism.

[0071] First, set up a Bloom filter. A Bloom filter is a very space-efficient probabilistic data structure used to determine whether an element belongs to a set. In this solution, the Bloom filter is used to track the use of cryptographic operations. The size of the Bloom filter m and the number of hash functions k need to be calculated based on the expected number of elements N and the acceptable false positive probability p. For example, assuming that 100,000 cryptographic operations are expected in an hour, and the false positive probability is expected to be no more than 0.1%, then it can be calculated that the size of the Bloom filter is about 1,437,759 bits, and 10 hash functions need to be used.

[0072] Next, a time decay function is introduced to calculate the key update weight. An exponential decay function is used here, where λ is the decay rate and t is the key usage time. For example, if λ=0.1 is set, then when the key is used for 10 time units, its weight will drop to about 36.8% of the initial value. This function ensures that newly used keys have higher weights, while the weights of keys that have not been used for a long time will gradually decrease.

[0073] Each time an encryption operation is performed, the system calculates a feature hash for that operation and inserts it into the Bloom filter. The feature hash can contain information such as the encryption timestamp, data type, user ID, etc. For example, an encryption operation performed by user A on a patient record at 10:00:00 on October 1, 2023, may generate a hash value of "2023100110000_userA_patientRecord".

[0074] The system periodically checks the fill rate of the Bloom filter. The fill rate refers to the proportion of bits in the Bloom filter that are set to 1. When this fill rate exceeds a preset fill threshold, the key update process is triggered. For example, the fill threshold can be set to 75%, which means that when more than 75% of the bits in the Bloom filter are set to 1, the key needs to be updated.

[0075] When a key update is triggered, the system generates a new key. This new key may be generated by a cryptographically secure random number generator. For example, a 256-bit AES key may be generated.

[0076] Then, the mixing weight is calculated using the exponential decay function defined above. This mixing weight determines the ratio of the new key to the old key in the final key. For example, if the old key has been used for 5 time units, its weight may be exp(-0.1*5) ≈ 0.61, and the weight of the new key is 1-0.61 = 0.39.

[0077] Based on these weights, the system mixes the new key with the old key to obtain the final updated dynamic encryption key. The mixing process may involve performing a bit-level XOR operation on the old key and the new key according to the weights.

[0078] Finally, the original data of each party is re-encrypted using this updated dynamic encryption key. This process may take quite a long time, depending on the amount of data. For example, for a dataset containing 1 million patient records, re-encryption may take several hours.

[0079] This process forms a complete closed loop: from setting up the Bloom filter, to tracking encryption operations, to triggering key updates, and finally generating and applying new keys. Each step is closely linked to the previous and next steps, together forming a dynamic and secure key management mechanism.

[0080] In this way, the system can adapt to changes in data usage patterns while maintaining high security. Frequently used data will trigger more frequent key updates, while infrequently used data may use relatively stable keys. This not only improves the security of the system, but also optimizes the use of computing resources because all data does not need to be rekeyed with the same frequency.

[0081] In general, this dynamic key update mechanism based on Bloom filter and time decay function provides a flexible, efficient and secure data protection solution for the collaborative learning model, effectively balancing security, performance and resource utilization.

[0082] Figure 2 FIG. 1 is a schematic diagram of the structure of a privacy data analysis system based on collaborative learning and dynamic encryption according to an embodiment of the present invention. Figure 2 As shown, the system comprises: The first unit is used to receive original data of multiple participants, where the original data of each participant is in an initial encryption state, and the initial encryption state is implemented by a homomorphic encryption algorithm; a unique participant identifier is assigned to each participant, and the participant identifier includes a timestamp and a random hash value; a corresponding dynamic encryption key is generated based on the participant identifier, and the dynamic encryption key is generated by a quantum key distribution technology; the original data of each participant is secondary encrypted by using the dynamic encryption key to obtain secondary encrypted data, and the secondary encryption adopts a layered encryption strategy, and encryption algorithms of different strengths are used for different sensitivity levels of data; The second unit is used to input the secondary encrypted data into a preset collaborative learning model, wherein the collaborative learning model includes multiple sub-models, each sub-model corresponds to a participant, and the sub-model adopts a federated learning architecture; the collaborative learning model is decrypted by using a secure computing protocol under homomorphic encryption and a dynamic encryption key of each participant to obtain a privacy protection analysis model that can be used by each participant; The third unit is used to process the input data to be analyzed by using the privacy protection analysis model during data analysis, and ensure the correctness of the analysis process through verifiable computing technology; based on the predefined access control policy, only the analysis results that the corresponding participants are entitled to access are returned, and the analysis results are k-anonymized to further enhance privacy protection; and the model update and retraining process are regularly executed to adapt to the dynamic changes in data distribution.

[0083] According to a third aspect of the embodiments of the present invention, An electronic device is provided, comprising: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0084] According to a fourth aspect of the embodiments of the present invention, A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the aforementioned method is implemented.

[0085] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0086] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A privacy data analysis method based on collaborative learning and dynamic encryption, characterized in that: include: Receive original data from multiple participants, where the original data of each participant is in an initial encryption state, and the initial encryption state is implemented using a homomorphic encryption algorithm; assign a unique participant identifier to each participant, and the participant identifier includes a timestamp and a random hash value; generate a corresponding dynamic encryption key based on the participant identifier, and the dynamic encryption key is generated using quantum key distribution technology; use the dynamic encryption key to perform secondary encryption on the original data of each participant to obtain secondary encrypted data, and the secondary encryption adopts a layered encryption strategy, and uses encryption algorithms of different strengths for different sensitivity levels of data; Inputting the secondary encrypted data into a preset collaborative learning model, wherein the collaborative learning model includes a plurality of sub-models, each sub-model corresponds to a participant, and the sub-model adopts a federated learning architecture; Using a secure computing protocol under homomorphic encryption and combining the dynamic encryption keys of each participant to decrypt the collaborative learning model, a privacy-preserving analysis model that can be used by each participant is obtained; When performing data analysis, the input data to be analyzed is processed using the privacy protection analysis model, and the correctness of the analysis process is ensured through verifiable computing technology; based on the predefined access control policy, only the analysis results that the corresponding participants are entitled to access are returned, and the analysis results are k-anonymized to further enhance privacy protection; The model update and retraining process is performed regularly to adapt to the dynamic changes in data distribution.

2. The method according to claim 1, characterized in that Generating a corresponding dynamic encryption key based on the participant identifier, wherein the dynamic encryption key is generated using quantum key distribution technology, includes: An entangled photon pair is generated by a periodically poled lithium niobate crystal through a spontaneous parametric down-conversion process, wherein the temperature of the periodically poled lithium niobate crystal is controlled at 163.5±0.1°C, and the wavelength of the entangled photon pair is 1550nm; one photon in the entangled photon pair is sent to a sender, and the other photon is sent to a receiver, wherein the photon transmission adopts a photonic crystal fiber with an optical loss rate not higher than 0.2dB / km; a quantum repeater is arranged every 50km on the transmission path of the photonic crystal fiber, wherein the quantum repeater adopts a rare earth-doped YSO crystal as a quantum storage medium, and the storage time is not less than 1ms; The sender and the receiver respectively use a superconducting nanowire single-photon detector to measure the quantum state of the received photons, wherein the quantum efficiency of the superconducting nanowire single-photon detector at a wavelength of 1550nm is not less than 93%, and the dark count rate is not higher than 1Hz; Based on the quantum state measurement result, a quantum key is generated through a post-processing step, which includes basis comparison, parameter estimation, information coordination and privacy amplification, wherein information coordination adopts a rate-adaptive low-density parity-check code, and privacy amplification adopts a privacy matrix with dynamically adjustable size.

3. The method according to claim 1, characterized in that The original data of each participant is re-encrypted using the dynamic encryption key to obtain re-encrypted data. The re-encryption adopts a layered encryption strategy, and encryption algorithms of different strengths are adopted for different sensitivity levels of data, including: A data sensitivity assessment model based on deep learning is used to perform sensitivity assessment on the encrypted data, wherein the data sensitivity assessment model based on deep learning includes 12 base layers, 2 fully connected layers, 1 Dropout layer and 1 output layer, wherein the dropout rate of the Dropout layer is 0.1, and the output layer uses a Sigmoid activation function; Based on the quantum key, three sub-keys are generated using a key derivation function, which are used for data encryption at three sensitivity levels: high, medium and low; For high-sensitivity data with a sensitivity score greater than 0.8, the Kyber1024 algorithm is used for key encapsulation, and AES-256-GCM is used for authenticated encryption; for medium-sensitivity data with a sensitivity score between 0.4 and 0.8, the NTRU-HRSS-701 parameter set is used for key generation and encapsulation, and ChaCha20-Poly1305 is used for authenticated encryption; for low-sensitivity data with a sensitivity score less than 0.4, the derived subkey is directly used, and XChaCha20-Poly1305 is used for authenticated encryption.

4. The method according to claim 1, characterized in that Before inputting the secondary encrypted data into the preset collaborative learning model, the method further includes training the collaborative learning model: During the training process of the collaborative learning model, each sub-model only processes the secondary encrypted data of the participant and generates an intermediate result, which includes the model parameter gradient and the local loss function value; a secure multi-party computing protocol is used to collect the intermediate results generated by each sub-model, and the intermediate results are aggregated to form a global model, and the aggregation process adopts differential privacy technology to prevent the reverse deduction of the original data; the global model is used to perform analysis and processing on the newly input data to be analyzed under privacy protection, and the analysis and processing includes feature extraction, classification, regression and clustering; during the training process of the collaborative learning model, the dynamic encryption key of each participant is periodically updated based on the Bloom filter and the time decay function, and the original data of each participant is re-encrypted using the updated dynamic encryption key; Monitoring the training process of the collaborative learning model, and triggering a model verification procedure when it is detected that the model performance reaches a preset threshold or a potential model poisoning attack is detected; The model verification procedure includes: verifying the integrity of the model using zero-knowledge proof technology, evaluating the robustness of the model using adversarial sample testing, and ensuring the fairness of the model through multi-party consistency checks; According to the model verification results, the collaborative learning strategy is dynamically adjusted, including updating the weights of participants, adjusting the learning rate, and modifying the model structure; when the model verification passes and the performance meets the requirements, the training process is terminated and the final collaborative learning model is output.

5. The method according to claim 4, characterized in that During the training process of the collaborative learning model, each sub-model processes only the secondary encrypted data of the participant and generates intermediate results, which include model parameter gradients and local loss function values; a secure multi-party computing protocol is used to collect the intermediate results generated by each sub-model, and the intermediate results are aggregated to form a global model. The aggregation process uses differential privacy technology to prevent reverse derivation of original data; The global model is used to perform analysis and processing of the newly input data to be analyzed under privacy protection, wherein the analysis and processing includes feature extraction, classification, regression and clustering, including: The original data of the participants is encrypted using the CKKS approximate homomorphic encryption scheme, wherein the polynomial modulus of the CKKS approximate homomorphic encryption scheme is 2^15, the plaintext modulus is 2^40, the ciphertext modulus is approximately 2^440, and the security level is greater than 128 bits; in the local sub-model of each participant, the encrypted data is batch encoded, multiple real numbers are packaged into a polynomial, and the nonlinear activation function is approximated using the Taylor expansion polynomial; forward propagation and back propagation calculations are performed on the encrypted data to generate encrypted model parameter gradients and encrypted local loss function values; Generate a public key and a private key using the (T,n) threshold Paillier encryption scheme, where T is the reconstruction threshold and n is the number of participants. Use the Shamir secret sharing scheme to split the private key into n parts and distribute them to each participant. Each participant uses the public key to encrypt its model parameter gradient and local loss function value, and adds calibrated Gaussian noise before encryption. The noise variance σ² of the calibrated Gaussian noise is calculated based on the preset privacy budget ε and sensitivity Δf. One participant is selected as an aggregator, and the other participants send the encrypted intermediate results to the aggregator, and the aggregator calculates the sum of the encrypted gradient and the encrypted loss; at least n participants cooperate to partially decrypt the encrypted total gradient and total loss using their respective private key shares, and the aggregator collects the partial decryption results and completes the final decryption to obtain the global gradient and global loss; The global model parameters are updated using the global gradient, and the global model is used to perform analysis and processing of newly input data to be analyzed under privacy protection; in the analysis and processing, the SimHash algorithm is used to extract features and map high-dimensional feature vectors to low-dimensional binary space; an exponential mechanism is used to select the best splitting features to construct a privacy-preserving decision tree ensemble; inner product function encryption is used to realize privacy-preserving linear regression; and Laplace noise is introduced in each iteration of the K-means clustering algorithm to protect the coordinates of the cluster center.

6. The method according to claim 5, characterized in that During the training process of the collaborative learning model, the dynamic encryption key of each participant is periodically updated based on the Bloom filter and the time decay function, and the original data of each participant is re-encrypted using the updated dynamic encryption key, including: Set a Bloom filter whose size m and number of hash functions k are calculated based on the expected number of elements N and the false positive probability p; use the exponential decay function w(t) = exp(-λt) to calculate the key update weight, where λ is the decay rate and t is the key usage time; For each encryption operation, its characteristic hash is calculated and inserted into the Bloom filter, and the filling rate of the Bloom filter is periodically checked; when the filling rate exceeds a preset filling threshold, a key update is triggered, a new key is generated, and the exponential decay function is used to calculate the mixing weight, the new and old keys are mixed to obtain the final key, and the updated dynamic encryption key is used to re-encrypt the original data of each participant.

7. A privacy data analysis system based on collaborative learning and dynamic encryption, used to implement the method described in any one of claims 1 to 6, characterized in that: include: The first unit is used to receive original data of multiple participants, where the original data of each participant is in an initial encryption state, and the initial encryption state is implemented by a homomorphic encryption algorithm; a unique participant identifier is assigned to each participant, and the participant identifier includes a timestamp and a random hash value; a corresponding dynamic encryption key is generated based on the participant identifier, and the dynamic encryption key is generated by a quantum key distribution technology; the original data of each participant is re-encrypted by using the dynamic encryption key to obtain re-encrypted data, and the re-encryption adopts a layered encryption strategy, and encryption algorithms of different strengths are used for different sensitivity levels of data; A second unit is used to input the secondary encrypted data into a preset collaborative learning model, wherein the collaborative learning model includes a plurality of sub-models, each sub-model corresponds to a participant, and the sub-model adopts a federated learning architecture; Using a secure computing protocol under homomorphic encryption and combining the dynamic encryption keys of each participant to decrypt the collaborative learning model, a privacy-preserving analysis model that can be used by each participant is obtained; The third unit is used to process the input data to be analyzed by using the privacy protection analysis model when performing data analysis, and ensure the correctness of the analysis process through verifiable computing technology; based on the predefined access control policy, only the analysis results that the corresponding participants are entitled to access are returned, and the analysis results are k-anonymized to further enhance privacy protection; The model update and retraining process is performed regularly to adapt to the dynamic changes in data distribution.

8. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images