Privacy computing system and method, electronic equipment and medium

Abstract

The invention provides a privacy calculation system and method, electronic equipment and a medium, and the system comprises a data obtaining module which is used for obtaining data calculation requests submitted by a plurality of participants and uploaded to-be-processed data; the data encryption module is used for performing encryption processing on the to-be-processed data based on a preset encryption algorithm; wherein the encryption algorithm comprises one of the following: an improved homomorphic encryption algorithm, a differential privacy algorithm, a zero-knowledge proof algorithm and federal learning; and the data calculation module is used for calculating the encrypted to-be-processed data based on the data calculation request to obtain a calculation result, and outputting the calculation result to the plurality of participants, so that the plurality of participants decrypt the calculation result to obtain the calculation result. According to the method, the data privacy is improved, and the risk of data leakage is reduced.

Description

Technical Field

[0001] The present invention relates to the field of privacy computing technology, and in particular to a privacy computing system, method, electronic device and medium. Background Art

[0002] Privacy computing is a technology that securely calculates agreed functions without sharing data between participants and without a trusted third party. A privacy computing platform is a system designed specifically to process and protect private information. With the increasing awareness of data security and privacy protection, the demand for privacy computing platforms is also growing. At present, existing privacy computing platform system products have achieved certain results in privacy protection and data security, but there are still some shortcomings and defects. Summary of the invention

[0003] In view of this, an object of the present invention is to provide a system, method, electronic device and medium for privacy computing, so as to improve data privacy and reduce the risk of data leakage.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0005] In the first aspect, the present invention provides a privacy computing system, comprising: a data acquisition module, used to obtain data computing requests submitted by multiple participants and uploaded data to be processed; a data encryption module, used to encrypt the data to be processed based on a preset encryption algorithm; wherein the encryption algorithm includes one of the following: an improved homomorphic encryption algorithm, a differential privacy algorithm, a zero-knowledge proof algorithm and federated learning; a data computing module, used to calculate the encrypted data to be processed based on the data computing request to obtain a calculation result, and output the calculation result to multiple participants, so that multiple participants can decrypt the calculation result to obtain the calculation result.

[0006] Optionally, the data encryption module is specifically used to: use an improved Harris Eagle optimization algorithm to randomly generate two prime numbers, and generate a key based on the two prime numbers; generate random numbers in a uniformly distributed manner, and encrypt the data to be processed based on the random numbers and the key; wherein the encryption process uses GPU parallel technology; and use a square multiplication algorithm and a Montgomery algorithm to calculate the encrypted data to be processed.

[0007] Optionally, the data calculation module is specifically used to: determine a target calculation function based on a data calculation request; and calculate the encrypted data to be processed based on the target calculation function to obtain a calculation result.

[0008] Optionally, it also includes: a federal preprocessing module, which is used to: preprocess the data to be processed based on a preset federal feature preprocessing operator; wherein the preprocessing at least includes: feature scale transformation, federal string indexing, federal missing value filling and federal standardization; and construct a federated learning model based on the preprocessed data to be processed.

[0009] Optionally, it also includes: a storage module, which is used to: segment and compress the data to be processed, and distribute the processed data to be processed.

[0010] Optionally, it also includes: a user authority module, which is used to: obtain the identity information of multiple participants, and perform identity authentication and access permission verification based on the identity information; if the verification is successful, obtain the data calculation requests submitted by multiple participants and the uploaded data to be processed.

[0011] Optionally, it also includes: a log recording module, which is used to: obtain log information during the calculation process, and detect in real time whether there are any abnormalities in the calculation process based on the log information. If so, an abnormality warning is issued; generate an operation log based on the log information, and audit and trace the calculation process based on the operation log.

[0012] In the second aspect, the present invention provides a method for privacy computing, which is applied to the privacy computing system of any one of the items provided in the first aspect, including: obtaining data computing requests submitted by multiple participants and uploaded data to be processed; encrypting the data to be processed based on a preset encryption algorithm; wherein the encryption algorithm includes one of the following: an improved homomorphic encryption algorithm, a differential privacy algorithm, a zero-knowledge proof algorithm and federated learning; calculating the encrypted data to be processed based on the data computing request to obtain a calculation result, and outputting the calculation result to multiple participants, so that the multiple participants can decrypt the calculation result to obtain the calculation result.

[0013] In a third aspect, the present invention provides an electronic device comprising a processor and a memory, wherein the memory stores computer executable instructions that can be executed by the processor, and the processor executes the computer executable instructions to implement the steps of the method provided in the second aspect above.

[0014] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the method provided in the second aspect are executed.

[0015] The present invention brings the following beneficial effects:

[0016] The above-mentioned privacy computing system, method, electronic device and medium provided by the present invention include: a data acquisition module, which is used to obtain data computing requests submitted by multiple participants and uploaded data to be processed; a data encryption module, which is used to encrypt the data to be processed based on a preset encryption algorithm; wherein the encryption algorithm includes one of the following: an improved homomorphic encryption algorithm, a differential privacy algorithm, a zero-knowledge proof algorithm and federated learning; a data computing module, which is used to calculate the encrypted data to be processed based on the data computing request to obtain a calculation result, and output the calculation result to multiple participants, so that multiple participants can decrypt the calculation result to obtain the calculation result. The above-mentioned system encrypts the data to be processed by means of an improved homomorphic encryption algorithm, a differential privacy algorithm, a zero-knowledge proof algorithm and federated learning, ensuring that data calculation and analysis are performed without exposing the original data, thereby improving data privacy and reducing the risk of data leakage.

[0017] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0018] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the preferred embodiments are specifically mentioned below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 A schematic diagram of the structure of a privacy computing system provided by an embodiment of the present invention;

[0021] Figure 2 A schematic diagram of the structure of another privacy computing system provided by an embodiment of the present invention;

[0022] Figure 3 A flowchart of a privacy computing method provided by an embodiment of the present invention;

[0023] Figure 4 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are 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.

[0025] At present, existing privacy computing platform system products have achieved certain results in privacy protection and data security, but there are still some shortcomings and defects:

[0026] (1) Computational efficiency needs to be improved: Although the privacy computing platform system uses advanced cryptographic techniques and algorithms to ensure data security, these technologies are often accompanied by high computing costs. In some complex computing scenarios, the existing privacy computing platform may not be able to meet the needs of efficient processing, resulting in slow computing speed and affecting the efficiency of practical applications.

[0027] (2) Performance bottleneck: Privacy computing usually involves complex encryption and computing processes, which may lead to degraded computing performance. Existing systems may encounter problems such as slow computing speed or excessive resource consumption when processing large-scale data sets, thus limiting their application in real-time or high-concurrency scenarios.

[0028] (3) High technical threshold: The cutting-edge technologies involved in privacy computing are relatively complex and require a high technical threshold for users. This may make it difficult for some non-professional users to understand and use these systems, limiting the popularity and application scope of the products.

[0029] (4) Security and trust issues: Although privacy computing platforms are designed to protect data privacy and security, in actual applications, users may still have doubts about the security of the platform. This includes concerns about data encryption strength, access control, audit tracking, etc., requiring the platform to provide sufficient security measures and trust mechanisms.

[0030] Based on this, the embodiments of the present invention provide a privacy computing system, method, electronic device and medium, which improve data privacy and reduce the risk of data leakage.

[0031] To facilitate understanding of this embodiment, a privacy computing system disclosed in an embodiment of the present invention is first introduced in detail. The system can be applied to multiple fields such as finance, insurance, medical care, education, and government affairs. First of all, the privacy computing system needs to be initialized and configured before use. This includes setting the basic parameters of the system, security policies (such as data encryption algorithms, computing models, security protocols, etc.), data formats, etc. In addition, it is also necessary to configure relevant computing resources and storage resources to ensure that the platform can operate normally.

[0032] See also Figure 1 The structural diagram of a privacy computing system shown in the figure illustrates that the system mainly includes the following parts: a data acquisition module 101, a data encryption module 102, and a data calculation module 103.

[0033] Among them, the data acquisition module is used to obtain data calculation requests submitted by multiple participants and uploaded data to be processed.

[0034] In one implementation, the participant (i.e., data provider) can upload the original data to be processed (i.e., data to be processed) and submit a data calculation request through the system interface or API interface. The data calculation request includes the calculation content required by the participant, i.e., what calculation method to use, what functions to use, etc. The participant inputs the data to be calculated into the privacy computing system, and the data can be pre-processed, including data cleaning, format conversion, etc., to ensure the security and availability of the data.

[0035] The data encryption module 102 is used to encrypt the data to be processed based on a preset encryption algorithm; wherein the encryption algorithm includes one of the following: an improved homomorphic encryption algorithm, a differential privacy algorithm, a zero-knowledge proof algorithm, and federated learning.

[0036] In one embodiment, the data encryption module 102 can apply privacy protection technologies such as differential privacy, federated learning, improved homomorphic encryption, zero-knowledge proof, etc. according to the configured security protocol to encrypt or convert the data to be processed to ensure that the original data is not leaked during the calculation process.

[0037] The data calculation module 103 is used to calculate the encrypted data to be processed based on the data calculation request to obtain a calculation result, and output the calculation result to multiple participants so that the multiple participants can decrypt the calculation result to obtain the calculation result.

[0038] In one implementation, the data computing module 103 may use multi-party secure computing technology to allow multiple parties to perform collaborative computing without sharing original data, including:

[0039] (1) Three-party private intersection function: Its basic principle is to use cryptographic technology to ensure that no party's data is leaked during the calculation process. Specifically, the three-party private intersection technology uses encryption algorithms and protocol design to enable the participating parties to calculate the intersection of data sets without sharing their respective data. Private intersection supports initial screening, complementation, or operations with calculation conditions, and performs multi-party secure calculations based on the three-party intersection results.

[0040] (2) Bucket screening and Bloom filter accelerate sample alignment by excluding non-overlapping IDs, ensuring excellent performance in the case of large amounts of data.

[0041] (3) Specify the privacy intersection of feature value range, label value, and sample time range.

[0042] Optionally, the data calculation module 103 is specifically used to: determine a target calculation function based on a data calculation request; and calculate the encrypted data to be processed based on the target calculation function to obtain a calculation result.

[0043] In specific implementation, the data calculation module 103 may first perform the calculation method required by the participant according to the data calculation request, that is, determine the target calculation function; and then use the encrypted data to be processed to calculate and obtain the calculation result.

[0044] In the embodiment of the present invention, the agreed function calculation can be completed securely without a trusted third party and without the participants sharing their respective private data in plain text, and the following multiple basic algorithms are supported:

[0045] (1) Support basic operators such as addition, subtraction, multiplication, division, and exponentiation;

[0046] (2) Support comparison operations such as less than, greater than, equal to, not equal to, greater than or equal to, less than or equal to;

[0047] (3) Supports multiple mathematical functions such as exponential function, logarithmic function, absolute value function, square root function, sign function, etc.

[0048] (4) Supports various trigonometric functions including sine function, cosine function, tangent function, hyperbolic sine function, hyperbolic cosine function, hyperbolic tangent function, etc.

[0049] (5) Supports various statistical functions such as sum, minimum, maximum, average, variance, etc.;

[0050] (6) A vertical logistic regression algorithm without a coordinator improves the adaptability in the two-party algorithm scenario;

[0051] (7) Vertical decision tree, Poisson regression, score card, K-Means algorithm.

[0052] After the calculation is completed, the calculation results will be generated and decrypted or de-perturbed as needed. The calculation results can be shared with participants in encrypted form or after desensitization, ensuring that the privacy of the data is maintained during the sharing process.

[0053] The above-mentioned privacy computing system provided by the present invention encrypts the data to be processed through improved homomorphic encryption algorithm, differential privacy algorithm, zero-knowledge proof algorithm and federated learning, ensuring data calculation and analysis without exposing the original data, thereby improving data privacy and reducing the risk of data leakage.

[0054] In one embodiment, the data encryption module 102 may use an improved homomorphic encryption algorithm to perform data encryption. The Paillier homomorphic encryption algorithm is a classic homomorphic encryption algorithm that can perform homomorphic encryption operations on encrypted data to achieve the effect of protecting data privacy. In order to ensure that data encryption has sufficient security, a 32-bit or 64-bit decimal can be converted into a large number with the same length as the key length (such as a large number with a length of 1024 bits or 2048 bits). Currently, a key length of 1024 bits or more is generally used.

[0055] The core idea of ​​the Pailier homomorphic encryption algorithm is to encrypt the plaintext twice with RSA, and then perform multiplication on the encrypted ciphertext. Specifically, the Pailier homomorphic encryption algorithm generates a public-private key pair, where the public key part includes the product of two prime numbers p and q, the L function, and the public key. When encrypting, first encrypt the plaintext m twice with the public key to obtain m1 and m2 respectively, then multiply m1 and m2, and then multiply them with the random number r to obtain the ciphertext c. When decrypting, the private key is needed to decrypt the ciphertext c. The decryption process is to divide the ciphertext c by m2^L modn to obtain E, and then use E to calculate m.

[0056] The specific steps are as follows:

[0057] 1. Key Generation

[0058] 1.1 Randomly select two prime numbers p and q, satisfying gcd(pq,(p-1)(q-1))=1, gcd is the greatest common divisor of the two numbers, and p and q are of equal length.

[0059] 1.2 Calculate n=pq and =lcm(p-1,q-1), where lcm represents the least common multiple.

[0060] 1.3 Randomly select integers (You can also set g=n+1 to optimize the calculation speed).

[0061] 1.4 Define L function: Calculate μ = (L (g λ modn 2 )) -1 modn.

[0062] After the above four steps, the public key: (n, g) and the private key: (, μ) are generated.

[0063] 2. Encryption

[0064] 2.1 Input the plaintext message m, satisfying 0 ≤ m ≤ n.

[0065] 2.2 Select a random number satisfying 0 < r < n and gcd=(r, n)=1, where gcd is the greatest common divisor of two numbers.

[0066] 2.3 Calculate the ciphertext: c = g m r n mod n 2 .

[0067] Among them, if g = n + 1, according to the binomial theorem: The first m - 1 terms are all multiples of n 2 and are eliminated under modulo n 2 Therefore, here the modular exponentiation operation is simplified to 1 modular multiplication operation, accelerating the calculation process, that is: g m = mn + 1 mod n 2 .

[0068] 3. Decryption

[0069] 3.1 Input the ciphertext c, satisfying

[0070] 3.2 Calculate the plaintext message m = L(c λ mod n 2 )·μ mod n.

[0071] Furthermore, the correctness can be verified through homomorphic addition and homomorphic scalar multiplication.

[0072] Correctness verification:

[0073] (1) Decryption correctness

[0074] According to the definition of the key, it can be deduced that when : g λ = 1 (mod n), g nλ = 1 (mod n 2 ).

[0075] Based on this, combined with the binomial theorem, the following derivation can be obtained:

[0076] d(c) = L(c λ mod n 2 )·μ mod n

[0077] = L((g m r n) λ modn 2 )·(L(g λ modn 2 )) -1 modn

[0078] =L((g λ ) m modn 2 )·(L(g λ modn 2 )) -1

[0079] =L((1+kn) m modn 2 )·(L(1+kn)) -1

[0080] =mk·k -1

[0081] =m

[0082] (2) Homomorphic addition correctness, that is, the addition of plaintext is a multiplication operation in the encryption domain.

[0083]

[0084] (3) Correctness of Homomorphic Scalar Multiplication

[0085]

[0086] Security verification: In terms of security, the Pailier homomorphic encryption algorithm can achieve a semi-blind state, that is, the numerator and denominator of the plaintext cannot be directly obtained by performing a division operation on the ciphertext. The Pailier homomorphic encryption algorithm introduces additional random numbers to perturb the interference results, thereby enhancing the security of the system. In addition, the Pailier homomorphic encryption algorithm can resist various attack methods such as chosen plaintext attacks, ciphertext attacks, and spectrum analysis attacks, and has high security performance.

[0087] The Pailier homomorphic encryption algorithm can achieve additive homomorphism and multiplicative homomorphism, has good data security performance, and can solve the problem of data privacy protection. However, since it involves multiple encryption operations, it will lead to low encryption and decryption efficiency, which greatly affects its application scope. Based on this, the embodiments of the present invention can be optimized in terms of parameter selection, GPU parallel acceleration, etc. to improve the efficiency of the Paillier homomorphic encryption algorithm.

[0088] Based on this, the improved Pailier homomorphic encryption algorithm is used to encrypt data in the embodiment of the invention, and the data encryption module 102 is specifically used for:

[0089] Firstly, the improved Harris Hawk optimization algorithm is used to randomly generate two prime numbers, and a key is generated based on the two prime numbers.

[0090] In the specific implementation, the improved Harris Hawks Optimization (HHO) algorithm can be used to select appropriate parameters (i.e. prime numbers p and q). Specifically, by introducing the elite hierarchy strategy, strengthening the communication between populations, making full use of the dominant population to estimate the better evolutionary direction of the population, and enhancing the diversity of the algorithm population; using Tent mapping to adjust the algorithm parameters; introducing a new update formula for the iteration of the energy factor E to balance the exploration and development capabilities of the algorithm; using the Gaussian random walk strategy for the optimal individual to increase the algorithm search area, and when the algorithm stagnates, using the Gaussian random walk strategy to increase the diversity of the population individuals to help the algorithm jump out of the local optimum; finally, using the greedy strategy to fully retain the dominant individuals and accelerate convergence.

[0091] Among them, in order to improve its global search ability and avoid the reduction of population diversity in the later iteration period, an elite hierarchy system is introduced. Considering the strengthening of suboptimal solution information exchange during the iteration process, three optimal positions are selected to replace the optimal solution to guide the remaining individuals to follow:

[0092]

[0093] Among them, X jbest is the dominant individual in the current population, f(X jbest ) is the fitness value of the dominant individual in the current population.

[0094] Then, random numbers are generated in a uniformly distributed manner, and the data to be processed is encrypted based on the random numbers and the key; wherein the encryption process uses GPU parallel technology.

[0095] In the specific implementation, in order to improve the efficiency of the Pailier homomorphic encryption algorithm, when encrypting plaintext, the random number r can be generated in a uniformly distributed manner, thereby increasing the difficulty for attackers to crack the random number and improving the security of the algorithm. At the same time, GPU parallel acceleration technology is introduced to improve the encryption and decryption efficiency of the algorithm without affecting the security of the algorithm.

[0096] Finally, the square multiplication algorithm and Montgomery algorithm are used to calculate the encrypted data to be processed.

[0097] In the specific implementation, the Paillier homomorphic encryption algorithm and the operation in the secret state use a large number of modular exponentiation operations. Efficiently calculating modular exponentiation operations through heterogeneous computing is the core of improving computing efficiency. In this embodiment, the computational complexity can be optimized by the square multiplication algorithm first. The advantage of the square multiplication algorithm is that it can reduce the complexity, but it needs to do n modular operations. Then the modular multiplication calculation is efficiently completed by the Montgomery modular multiplication algorithm. The advantage of the Montgomery algorithm is that no modular operation is required in the process of calculating modular multiplication, thereby greatly speeding up the modular operation speed.

[0098] In addition, in an embodiment of the present invention, the Pailier homomorphic encryption algorithm can be used in combination with a deep learning algorithm, and the algorithm parameters can be optimized through machine learning methods to improve the performance of the algorithm.

[0099] See also Figure 2 As shown, the above system also includes: a federal preprocessing module 104, which is used to: preprocess the data to be processed based on a preset federal feature preprocessing operator; wherein the preprocessing at least includes: feature scale transformation, federal string indexing, federal missing value filling and federal standardization; and construct a federal learning model based on the preprocessed data to be processed.

[0100] In specific implementation, the above system provides a variety of federated feature preprocessing operators. When configuring a federated task, users can define federated feature preprocessing operators to perform feature processing on data. The preprocessing methods include: filtering, feature scale transformation, federated string indexing, federated missing value filling, and federated standardization. The preprocessing methods are described as follows:

[0101] (1) Filtering: This operator can filter data according to different conditions and provides a variety of methods to process data, such as filtering data containing missing values ​​or filtering data according to custom conditions. There are three types of filtering conditions: notMissing means filtering non-missing values; notMissingLabel means filtering non-missing labels; and notMissingAttributes means filtering non-missing attributes.

[0102] (2) Feature scaling: Perform scaling transformations such as log, abs, and sqrt on the input features.

[0103] (3) Federated string indexing: The function of a string index is to map k different strings to k integers between [0, k-1], thereby completing the transformation from strings to numbers. Federated string indexing is to find all the values ​​that appear in the target column in a federated learning scenario and map strings to numbers.

[0104] (4) Federal missing value filling: Federal missing value filling mainly includes two parts: calculating the global mean and using the global mean to fill in missing values.

[0105] (5) Federated normalization: Normalization is the process of calculating the mean μ and standard deviation σ of the target column and performing a (x-μ) / σ transformation on each element x of the column. The purpose of normalization is to make the processed data follow a standard normal distribution. The difference of federated normalization is that the mean μ and standard deviation σ are calculated using global data, not just local data.

[0106] Furthermore, the pre-processed data to be processed is used to build a federated learning model. In this embodiment, common statistical analysis and machine learning algorithms such as horizontal and vertical logistic regression, linear regression, and K-means are supported, and personalized privacy computing algorithms are adapted and transformed for a variety of scenarios and algorithms to address the problem of inconsistent data distribution in privacy computing. In addition, it also supports the output of rich data quality and algorithm model evaluation indicators to help users determine the business effects of federated data quality and models in the specified scenarios, and to achieve the measurement of the quality and importance of each party's features under the federation, and the effective quantification of each party's data contribution and model effect.

[0107] See also Figure 2 As shown, the above system also includes: a storage module 105, which is used to: perform sharding and data compression on the data to be processed, and perform distributed storage on the processed data to be processed.

[0108] In specific implementation, in order to reduce storage space occupation and network bandwidth requirements, in the embodiments of the present invention, after the data to be processed is acquired, the data to be processed may be segmented and the data may be stored through a distributed storage strategy.

[0109] See also Figure 2 As shown, the above system also includes: a user authority module 106, which is used to: obtain the identity information of multiple participants, and perform identity authentication and access permission verification based on the identity information; if the verification is successful, obtain the data calculation requests submitted by multiple participants and the uploaded data to be processed.

[0110] The above system provided in this embodiment provides a permission management and access control mechanism to ensure that only authorized users or organizations can access and use the data and functions on the platform. Specifically, unauthorized access and data leakage can be prevented through identity authentication, role management, access permission control and other means.

[0111] In specific implementation, after the participant logs into the system, the participant's identity information can be obtained, and identity authentication and access permission verification can be performed along with the identity information to verify whether the participant is an authorized user or organization. If the verification is successful, the participant is allowed to access the system, submit data calculation requests and upload data to be processed.

[0112] See also Figure 2 As shown, the above system also includes: a log recording module 107, which is used to: obtain log information during the calculation process, and detect in real time whether there is any abnormality in the calculation process based on the log information, and if so, issue an abnormality warning; generate an operation log based on the log information, and audit and trace the calculation process based on the operation log.

[0113] In the specific implementation, during the entire calculation process, the log recording module 107 will record detailed log information, including: data input, calculation process, result output, etc., and can detect abnormal situations or potential risks in the calculation process in real time based on the log information to ensure the safety and accuracy of the calculation, and issue abnormal warnings when abnormalities are detected to enhance the staff's relevant intervention processing.

[0114] At the same time, you can also generate operation logs to record key information in the calculation process for subsequent auditing and troubleshooting. By viewing the operation logs, users can trace the operators and execution status used in each step of the task execution process, which helps to enhance users' trust in the system and ensure the authenticity and validity of the calculation results. In addition, you can also connect to log auditing, traffic analysis, WAF and other systems, and then use the big data intelligent security platform for unified security operations to build an intelligent security operation protection system.

[0115] In the embodiments of the present invention, the privacy computing system can also be continuously optimized and updated, including improving algorithm efficiency, enhancing security performance, expanding functional scope, etc. Through continuous optimization and updating, the privacy computing system can better meet the needs of users and promote the further release and application of data value. Specifically, it can be optimized and upgraded regularly according to user feedback and actual needs to improve computing performance, security and user experience, and introduce new privacy protection technologies and algorithms to adapt to the ever-changing data privacy needs and security challenges.

[0116] The privacy computing system provided by the embodiment of the present invention is based on data security technologies such as homomorphic encryption, federated learning and differential privacy to achieve secure data utilization and analysis under privacy protection, ensuring that data is "available but invisible". Compared with existing privacy computing platforms, it has the following advantages:

[0117] (1) Algorithm optimization and technological innovation: By continuously optimizing cryptographic algorithms and privacy computing technologies, we can improve computing efficiency and reduce computing costs. At the same time, we can introduce new technical means, such as distributed computing and parallel processing, to further accelerate data processing speed and meet the needs of large-scale data processing.

[0118] (2) Resource management and optimization: To address the bottleneck issues of storage and transmission, the privacy computing system uses resource management technology, compression technology, and optimized network transmission protocols, such as data compression, distributed storage, encoding, and stream processing, to effectively reduce storage space usage and network bandwidth requirements. At the same time, through reasonable data sharding and distributed storage strategies, data transmission protocols are optimized, and network bandwidth utilization is improved, so that large-scale data can be better processed, data processing speed and efficiency can be improved, and rapid data transmission can be ensured.

[0119] (3) Simplify the operation process: The privacy computing platform system is committed to reducing the complexity of operations. By providing a more intuitive and friendly user interface and operation process, it reduces the user's learning cost and difficulty of use. At the same time, it provides detailed operation guides and technical support to help users quickly get started and become familiar with various functions.

[0120] (4) Unified interface and standardization: In order to solve the compatibility issues between different platforms, the privacy computing platform system actively promotes interface unification and standardization. By formulating unified technical standards and interface specifications, seamless integration between different platforms can be achieved, the compatibility and ease of use of the system can be improved, and the integration difficulty and cost of users can be reduced during use.

[0121] (5) Enhanced security mechanism: The privacy computing platform system adopts a multi-layer security protection mechanism, including identity authentication, access control, data encryption, etc., to effectively prevent hacker attacks and data leakage, and ensure the integrity and confidentiality of data. At the same time, the system is regularly inspected and vulnerabilities are repaired to respond to potential security threats in a timely manner.

[0122] (6) Upgrading privacy protection technology: By adopting more advanced differential privacy, homomorphic encryption, zero-knowledge proof and other technical means, we ensure that calculations and analysis are performed without exposing the original data, thereby ensuring the privacy of the data. In addition, we strengthen the trust management among the participants, introduce third-party supervision or blockchain and other technical means, and strictly manage the rights and access control mechanisms to ensure that only authorized users can access and operate the data, and that all parties comply with the agreement and maintain the privacy of the data.

[0123] Regarding the privacy computing posture provided by the aforementioned embodiment, the present invention also provides a privacy computing method, which is applied to the privacy computing system provided by the aforementioned embodiment. Figure 3The flowchart of the privacy computing method shows that the method mainly includes steps S301 to S303:

[0124] Step S301: obtaining data calculation requests submitted by multiple participants and uploaded data to be processed;

[0125] Step S302: Encrypt the data to be processed based on a preset encryption algorithm; wherein the encryption algorithm includes one of the following: an improved homomorphic encryption algorithm, a differential privacy algorithm, a zero-knowledge proof algorithm, and federated learning;

[0126] Step S303: Based on the data calculation request, the encrypted data to be processed is calculated to obtain a calculation result, and the calculation result is output to multiple participants, so that the multiple participants can decrypt the calculation result to obtain the calculation result.

[0127] The above-mentioned privacy computing method provided by the present invention encrypts the data to be processed through improved homomorphic encryption algorithm, differential privacy algorithm, zero-knowledge proof algorithm and federated learning, ensuring that data calculation and analysis are performed without exposing the original data, thereby improving data privacy and reducing the risk of data leakage.

[0128] It should be noted that the implementation principle and technical effects of the method provided in the embodiment of the present invention are the same as those of the aforementioned system embodiment. For the sake of brief description, for matters not mentioned in the method embodiment, reference may be made to the corresponding contents in the aforementioned system embodiment.

[0129] An embodiment of the present invention further provides an electronic device. Specifically, the electronic device includes a processor and a storage device. The storage device stores a computer program, and when the computer program is executed by the processor, it executes the method described in any one of the above implementation methods.

[0130] Figure 4 A structural diagram of an electronic device provided in an embodiment of the present invention, the electronic device 100 includes: a processor 40, a memory 41, a bus 42 and a communication interface 43, wherein the processor 40, the communication interface 43 and the memory 41 are connected via the bus 42; the processor 40 is used to execute an executable module stored in the memory 41, such as a computer program.

[0131] The memory 41 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 43 (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used.

[0132] The bus 42 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0133] Among them, the memory 41 is used to store programs, and the processor 40 executes the program after receiving the execution instruction. The method executed by the device for flow process definition disclosed in any embodiment of the above-mentioned embodiment of the present invention can be applied to the processor 40 or implemented by the processor 40.

[0134] The processor 40 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 40. The above processor 40 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in the embodiments of the present invention can be directly embodied as a hardware decoding processor to execute, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 41, and the processor 40 reads the information in the memory 41 and completes the steps of the above method in combination with its hardware.

[0135] The computer program product of the readable storage medium provided in the embodiment of the present invention includes a computer-readable storage medium storing program code, and the instructions included in the program code can be used to execute the methods described in the previous method embodiments. The specific implementation can be referred to the previous method embodiments, which will not be repeated here.

[0136] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0137] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-mentioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A privacy computing system, characterized in that: include: The data acquisition module is used to obtain data calculation requests submitted by multiple participants and uploaded data to be processed; A data encryption module, used to encrypt the data to be processed based on a preset encryption algorithm; wherein the encryption algorithm includes one of the following: an improved homomorphic encryption algorithm, a differential privacy algorithm, a zero-knowledge proof algorithm, and federated learning; The data calculation module is used to calculate the encrypted data to be processed based on the data calculation request to obtain a calculation result, and output the calculation result to the multiple participants so that the multiple participants can decrypt the calculation result to obtain the calculation result.

2. The system according to claim 1, characterized in that The data encryption module is specifically used for: Using an improved Harris Eagle optimization algorithm to randomly generate two prime numbers, and generating a key based on the two prime numbers; Generate a random number in a uniformly distributed manner, and encrypt the data to be processed based on the random number and the key; wherein the encryption process adopts GPU parallel technology; The encrypted data to be processed are calculated using the square multiplication algorithm and the Montgomery algorithm.

3. The system according to claim 1, characterized in that The data calculation module is specifically used for: Determining a target computing function based on the data computing request; The encrypted data to be processed is calculated based on the target calculation function to obtain a calculation result.

4. The system according to claim 1, characterized in that Also includes: Federated preprocessing module for: Preprocessing the data to be processed based on a preset federated feature preprocessing operator; wherein the preprocessing at least includes: feature scale transformation, federated string indexing, federated missing value filling and federated standardization; Build a federated learning model based on the preprocessed data to be processed.

5. The system according to claim 1, characterized in that Also includes: Storage modules for: The data to be processed is subjected to sharding and data compression processing, and the processed data to be processed is distributedly stored.

6. The system according to claim 1, characterized in that Also includes: User permissions module, used to: Obtaining identity information of the multiple participants, and performing identity authentication and access rights verification based on the identity information; If the verification is successful, the data calculation requests submitted by multiple participants and the uploaded data to be processed are obtained.

7. The system according to claim 1, characterized in that Also includes: Logging modules for: Obtain log information during the calculation process, and detect in real time whether there is any abnormality in the calculation process based on the log information. If so, issue an abnormality warning; An operation log is generated based on the log information, and the computing process is audited and traced based on the operation log.

8. A privacy computing method, characterized in that: A privacy computing system applied to any one of claims 1 to 7, comprising: Obtain data calculation requests submitted by multiple participants and uploaded data to be processed; Encrypting the data to be processed based on a preset encryption algorithm; wherein the encryption algorithm includes one of the following: an improved homomorphic encryption algorithm, a differential privacy algorithm, a zero-knowledge proof algorithm, and federated learning; The encrypted data to be processed is calculated based on the data calculation request to obtain a calculation result, and the calculation result is output to the multiple participants so that the multiple participants decrypt the calculation result to obtain the calculation result.

9. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores computer executable instructions that can be executed by the processor, and the processor executes the computer executable instructions to implement the steps of the method according to claim 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to claim 8 are performed.

Images