A privacy protection method for multi-party data, application and device thereof

Abstract

This invention discloses a method and its application / device for privacy protection of multi-party data. The method includes: S1: Selecting a cyclic group based on the number of participants to generate group element sequences in different orders; S2: Each participant generates a random number and a blind factor. The random number hides the participant's original data, and the blind factor hides the random number; S3: Accumulating and passing the sum of the random number and the blind factor in different orders of the group element sequences to obtain a first accumulation result, and accumulating and passing the blind factor in different orders of the group element sequences to obtain a second accumulation result; S4: Subtracting the second accumulation result from the first accumulation result to obtain the sum of random numbers; S5: Taking the opposite of the sum of random numbers as the random number of the last participant; S6: Accumulating and passing the random number with the original data to obtain the sum of the original data. This invention can effectively avoid collusion attacks by malicious nodes, improve the security and flexibility of multi-party data summation, and achieve privacy protection of the original data.

Description

Technical Field

[0001] This invention relates to the technical field of data security and privacy protection, and in particular to a method and device for protecting the privacy of multi-party data. Background Technology

[0002] Data security and privacy protection are major issues facing the field of data application today. With the rapid development of technologies such as big data and artificial intelligence, data has become one of the core resources of the world today, but it has also raised increasing concerns and challenges. Data leaks, misuse, or improper handling can have serious impacts on individuals, organizations, and even national security.

[0003] Secure multi-party summation, a fundamental protocol in cryptography, enables the summation of data from multiple parties without revealing the original data, ensuring the security of all parties' data. This technology is of great significance in the current field of data security and privacy protection, and can be applied to data analysis and value sharing in scenarios such as healthcare, finance, and social media.

[0004] Currently available technologies typically employ two methods: One method involves randomizing the original data by adding a random number. A master node is selected from multiple participants. This node generates a random number, adds its own data to it, and sends it to the next participant. Subsequent participants continue this process, accumulating the sum until it reaches the master node. The master node then subtracts the random number to obtain the sum of the data. The drawback of this method is its simplistic transmission order and weak security; collusion among adjacent participants can allow them to obtain the data of intermediate participants.

[0005] Another approach is based on segmenting the original data and broadcasting or transmitting the segmented data. Depending on the transmission order, this can be further divided into fixed-order transmission and variable-order transmission. The drawbacks of this method are that fixed-order transmission generally requires public-key encryption, involving numerous encryption and decryption operations and incurring significant computational overhead; variable-order transmission has overly simplistic rules, making it difficult to effectively address collusion between adjacent nodes, resulting in weaker security.

[0006] In addition, existing technical solutions also have the problem of not being able to flexibly configure security policies according to specific application scenarios and security levels. Summary of the Invention

[0007] The purpose of this invention is to address the technical problems of weak security and insufficient flexibility in existing multi-party data summation methods, and to propose a privacy protection method for multi-party data, as well as its application and device.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: A method for protecting the privacy of multi-party data includes the following steps: S1: Selecting a cyclic group based on the number of participants to generate group element sequences of different orders; S2: Each participant generates a random number and a blind factor, wherein the random number is used to hide the participant's original data, and the blind factor is used to hide the random number; S3: Accumulating and passing the sum of the random number and the blind factor according to different orders of the group element sequences to obtain a first accumulation result, and accumulating and passing the blind factor according to different orders of the group element sequences to obtain a second accumulation result; S4: Subtracting the second accumulation result from the first accumulation result to obtain the sum of the random numbers; S5: Taking the opposite of the sum of the random numbers as the random number of the last participant; S6: Accumulating and passing the random number with the original data to obtain the sum of the original data of the participants, thereby protecting the privacy of the original data.

[0009] In some embodiments of the present invention, an additive cyclic group is selected based on the number of participants, wherein the number of group elements in the additive cyclic group is greater than or equal to the number of participants, and each participant corresponds to one group element; the group element sequence is regenerated based on the different generators of the cyclic group.

[0010] In some embodiments of the present invention, the generator of the additive cyclic group is a set S, which consists of integers less than n and relatively prime to n, where n is a prime number greater than or equal to the number of participants; if the number of participants is not a prime number, then n is a prime number greater than the number of participants; in different sequences of group elements, the next adjacent group element of the same group element is different; the order of the group elements regenerated by different generators is different; the last group element of the sequences of group elements regenerated by different generators is the same.

[0011] In some embodiments of the present invention, step S3 includes the following steps: S31: Each participant adds its own random number to the blind factor, and the result is accumulated and passed on sequentially according to the order of the group elements regenerated by different generators. After all random numbers are added to the blind factor, the first accumulated result is obtained; S32: Each participant accumulates and passes on its own blind factor according to the order of the group elements regenerated by different generators. After all blind factors are added, the second accumulated result is obtained.

[0012] In some embodiments of the present invention, the last group element of the group element sequence corresponds to a determined participant, and the sum of the random numbers is calculated by the determined participant.

[0013] In some embodiments of the present invention, the participating party adds its own original data to the random number to obtain ciphertext data, and broadcasts the ciphertext data to all participating parties in a broadcast mode. Each participating party adds the ciphertext data received from other participating parties to its own ciphertext data to obtain the sum of the original data, so that each participating party can obtain the sum of the original data.

[0014] In some embodiments of the present invention, the participant adds its original data to the random number to obtain ciphertext data, and passes the ciphertext data to the next participant in a sequential transmission mode, so that only one participant receives the sum of the original data of the participants.

[0015] This invention also provides a joint credit investigation method, comprising the following steps: A1: obtaining a credit investigation request from a user to be investigated; A2: selecting at least three parties from the following: a bank, a tax authority, a public security authority, and the user's workplace; A3: using the privacy protection method for multi-party data as described in any of the preceding steps to obtain a sum of random numbers, and making the random number of the last participating party the opposite of the sum of the random numbers; A4: each participating party standardizes the original data, and the standardized data corresponds to the credit score of the user to be investigated; A5: each participating party adds the random number to the standardized data, and after accumulating and transmitting, obtains the standardized total credit score of the user to be investigated, thereby obtaining the joint credit investigation information of the user to be investigated.

[0016] This invention also provides a method for detecting abnormal financial behavior, comprising the following steps: B1: obtaining a financial abnormal behavior inspection instruction from the user to be detected; B2: selecting at least three financial institutions as participants; B3: using the privacy protection method for multi-party data as described above to obtain a sum of random numbers, and making the random number of the last participant the opposite of the sum of the random numbers; B4: each participant standardizes the original data, and the standardized data corresponds to the abnormal transaction amount of the user to be detected; B5: each financial institution adds the random number to the standardized data, and after accumulating and transmitting, obtains the standardized total abnormal transaction amount of the user to be detected, thereby detecting whether the user to be detected has potential abnormal financial behavior.

[0017] This invention also provides a method for sharing medical data, comprising the following steps: C1: obtaining statistical instructions from users, wherein the statistical instructions are one of the following for each medical institution: number of patients, case statistics, and disease incidence rate; C2: selecting at least three medical institutions as participants; C3: using the privacy protection method for multi-party data as described in any of the above steps to obtain a sum of random numbers, and making the random number of the last participant the opposite of the sum of the random numbers; C4: each participant standardizes the original data, and each participant adds the random number to the standardized data, and after accumulating and transmitting, obtains at least one of the following for each medical institution: number of patients, case statistics, and disease incidence rate.

[0018] The present invention also provides an apparatus including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the privacy protection method for multi-party data as described in any of the above.

[0019] The present invention has the following beneficial effects: The privacy protection method for multi-party data proposed in this invention first generates a sequence of group elements in different orders by selecting a cyclic group based on the number of participants. Then, it generates random numbers and blind factors. The random numbers are used to hide the original data of the participants, and the blind factors are used to hide the random numbers. This method can effectively avoid collusion attacks by malicious nodes and can select different numbers of blind factors according to different application scenarios and needs, thereby improving the security and flexibility of the multi-party data summation method and protecting the privacy of the original multi-party data.

[0020] In addition, some embodiments also have the following beneficial effects: By fixing the last participant in the group element sequence in each round, the computational efficiency of summing multi-party data can be improved.

[0021] Other beneficial effects of the embodiments of the present invention will be further described below. Attached Figure Description

[0022] Figure 1 This is a flowchart of the steps of the privacy protection method for multi-party data in an embodiment of the present invention; Figure 2 This is an example diagram illustrating the generation of random numbers and blinding factors by each participant in Example 1; Figure 3 This is an example diagram of generating different group element sequences in Example 1; Figure 4 This is a block diagram of the multi-party secure summation system proposed in the embodiments of the present invention. Detailed Implementation

[0023] The present invention will be further described below with reference to the accompanying drawings and preferred embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0024] It should be noted that the directional terms such as left, right, up, down, top, and bottom used in this embodiment are only relative concepts or are based on the normal use of the product, and should not be considered as restrictive.

[0025] To address the issues of weak security, high computational overhead, and insufficient flexibility in existing methods for summing multi-party data, this invention, based on the properties of cyclic groups, introduces cyclic groups into a secure multi-party summation method for the first time, proposing a privacy protection method for multi-party data, along with its application and device, which can achieve a balance between security and efficiency.

[0026] The following embodiments of the present invention propose a method for protecting the privacy of multi-party data, the steps of which are as follows: Figure 1 As shown, the explanation is as follows.

[0027] S1: Select a cyclic group based on the number of participants to generate group element sequences of different orders, in order to solve problems such as overly simple change rules and weak security caused by collusion between adjacent nodes.

[0028] According to the participants P quantity N , respectively Choose an appropriate addition cycle group ,in The identity element is 0, the number of group elements n is greater than or equal to the number of participants N, and each participant corresponds to a group element.

[0029] Determine the generators of the cyclic group g , generator g For a set ,gather From less than n And with n The group consists of coprime integers. When the number of participants is a prime number, the number of group elements n is equal to the number of participants; if the number of participants is not a prime number, then the number of group elements n is a prime number greater than and closest to the number of participants. Number of generators t Euler's totient function can be used It means, that is The group elements are regenerated from the generators, and the group elements can have different arrangements.

[0030] For example: If there are 6 participants, and 6 is not a prime number, then it's an addition cyclic group. The generators are only 1 and 5, which cannot generate a sufficient number of group element sequences, making it difficult to guarantee security. Therefore, the number of group elements n should be a prime number greater than 6 and closest to 6, namely 7. The generators are 1, 2, 3, 4, 5, and 6, which can satisfy the requirement of generating different sequences of group elements.

[0031] If the number of elements n in the group is equal to the number of participants N, then... All of them are actual participants.

[0032] If the number of elements n in the group is greater than the number of participants N, As actual participants, then It can be used as an additional virtual participant.

[0033] It has the following properties: (1) Any two generators The generated group element sequence ,have That is, the order of the group elements is different; and for the same group element ,have That is, in different group element sequences, the same element The next adjacent group element is different; (2) Arbitrary generator The generated group element sequence ,have That is, group The last element must be 0.

[0034] Based on the aforementioned properties of cyclic groups, the order of the group element sequences regenerated by different generators is different, thus ensuring security; the last element of the group sequences generated by different generators is the same, thus improving efficiency.

[0035] S2: Generate two sets of data, one set consisting of random numbers. One set of random numbers is generated independently locally by each participant to hide their original data; the other set of random numbers is a blinding factor. Each participant generates its own random number locally to hide it.

[0036]

[0037] Indicates the first The first participant generated the first One blind factor, The number of blinding factors generated for each participant, and In specific embodiments, different numbers of blind factors can be generated according to different security level requirements, with a higher number representing a higher security level.

[0038] S3: Data transfer, which involves accumulating and transferring the sum of the random number and the blind factor according to different orders of the group element sequence to obtain the first accumulation result, and accumulating and transferring the blind factor according to different orders of the group element sequence to obtain the second accumulation result. This includes three steps: Step 1, Select generator Generate a sequence of group elements The last element of the group element sequence corresponds to a specific participant (let's assume it's a participant). Each participant Add each random number to all blind factors, i.e. According to the group element sequence Different orders are passed on sequentially, and the last participant... The accumulated value held is Record this as the first accumulated result. .

[0039] Step 2, Select generator Generate a sequence of group elements Each participant will assign their respective blinding factor. According to the group element sequence The different orders are passed on sequentially. After one round, the last participant... The accumulated value held is .

[0040] Step three, repeat step two until all blinding factors have been passed. At this point, the last participant... The accumulated value held is the second accumulated result. .

[0041] S4: Subtract the second accumulation result from the first accumulation result to calculate the sum of random numbers. That is, by the participating party Calculate other The sum of random numbers from each participant The result is for all others The sum of random numbers from each participant .

[0042] S5: The last participant's random number is the negative of the sum of random numbers, that is, let... random numbers Then all The random numbers from each participant can satisfy... And none of them know the random numbers generated by the other participants. Let The purpose is to use random numbers Hide raw data This will not affect the final summation result of all original data.

[0043] S6: The random number is accumulated and transmitted with the original data to calculate the sum of the original data of the participants. Each participant Use your original data With random numbers Adding them together yields the encrypted data. The encrypted data is then transmitted sequentially to the next participant either in broadcast mode (i.e., all participants receive the message) or in order mode (i.e., only the next participant receives the message), resulting in the following final outcome: ,because Then you can get The result is as follows. In broadcast mode, each participant can obtain the sum of the original data by adding the ciphertext data received from other participants to their own ciphertext data (i.e., a secure summation result). In a sequential transmission mode, each participant passes their encrypted data to the next participant, ensuring that only the last participant receives the sum of the participants' original data (i.e., the secure summation result). During this process, no information is leaked. The original value is obtained, thereby protecting the privacy of the original data.

[0044] Symbol total: :all The sum of the original data of each participant, i.e. ; : The sum of random numbers from each participant, i.e. ; : The sum of the blinding factors of each participant, i.e. ; : The sum of the random numbers of each participant and the blind factor, i.e. .

[0045] In this embodiment of the invention, the order of the group element sequence is different in each round of the multi-party secure summation process using a cyclic group, which effectively avoids collusion attacks by malicious nodes. Furthermore, the last participant in each round of the group element sequence is fixed, improving efficiency. Simultaneously, more blind factors can better address the collusion problem; therefore, different numbers of blind factors can be generated and more rounds of transmission can be executed according to different security requirements, thereby flexibly selecting the security level and achieving a balance between security and efficiency.

[0046] Furthermore, existing technical solutions cannot flexibly configure security policies according to specific application scenarios and security levels. This makes it difficult to effectively address potential security threats in different contexts and fails to maximize the balance between security and efficiency. To apply the aforementioned privacy protection methods for multi-party data in different fields, embodiments of the present invention also include the following:

[0047] This invention also proposes a joint credit investigation method. When financial institutions investigate customer credit, they typically need to collect data from multiple parties, such as banks, tax authorities, public security organs, and the employer of the user to be investigated, for joint calculation. The method includes the following steps: A1: Obtain the credit information request of the user to be tested; A2: Select at least three entities from the following: banks, tax authorities, public security organs, and the workplace of the user to be tested, as participating parties; A3: Obtain the original sum of random numbers using any of the above-described privacy protection methods for multi-party data, and make the random number of the last participant the opposite of the sum of the random numbers, specifically including the following steps: A301: Select a cyclic group based on the number of participants and generate several different sequences of group elements; A302: Each participant generates a random number and several blind factors; A303: Add the random number to the blind factor, and pass the sums in the order of the group element sequence to obtain the first accumulated result; A304: The blind factor is accumulated and passed in the order of different group element sequences to obtain the second accumulation result; A305: Subtract the second accumulated result from the first accumulated result to obtain the sum of random numbers; A306: Let the random number of the last participant be the opposite of the sum of the above random numbers, so that the sum of the random numbers of all participants is 0; A4: Each participant will standardize the original data to achieve data additivity. The standardized data of each participant corresponds to the user's credit score for that item. A5: Each participant adds the random number to the standardized data, and accumulates and passes the data in the order of the group element sequence to obtain the sum of the standardized data, which corresponds to the user's total credit score; thereby realizing the joint credit investigation using the above-mentioned privacy protection method of multi-party data. This embodiment of the invention can complete the joint credit investigation without disclosing the specific values ​​of each party.

[0048] This invention also proposes a method for detecting abnormal financial behavior, comprising the following steps: B1: Obtain instructions for checking abnormal financial behavior of the user to be detected; B2: Select at least three financial institutions as participants and obtain the raw data of the users to be tested; B3: Using any of the above-described privacy protection methods for multi-party data, the sum of random numbers is obtained as the sum of the original data, and the random number of the last participant is the opposite of the sum of the original data. Specifically, this includes the following steps: B301: Select a cyclic group based on the number of participants, and generate several different sequences of group elements; B302: Each participant generates a random number and several blind factors; B303: Add the random number to the blind factor, and pass the sums in the order of the group element sequence to obtain the first accumulated result; B304: The blind factor is accumulated and passed in the order of different group element sequences to obtain the second accumulation result; B305: Subtract the second accumulation result from the first accumulation result to obtain the sum of random numbers; B306: Let the last participant's random number be the opposite of the sum of the above random numbers, so that the sum of the random numbers of all participants is 0; B4: Each participant will standardize the raw data to achieve data additivity. The standardized data of each participant corresponds to the abnormal transaction amount of the user to be detected. B5: Each participant adds the random number to the standardized data, and passes the sum in the order of the group element sequence to obtain the sum of the standardized data, which corresponds to the total abnormal transaction amount of the user to be detected; thereby realizing the detection of potential financial abnormal behavior of the user to be detected using the above-mentioned privacy protection method of multi-party data.

[0049] The embodiments of the present invention enable multiple institutions to collaboratively calculate the total abnormal transaction amount of a customer in order to detect potential abnormal behavior. In this process, it is not necessary to share the specific transaction details of the customer with each institution, thereby protecting the customer's privacy.

[0050] In the field of medical data statistical analysis, medical data involves a significant amount of sensitive personal information, and data sharing and application across medical institutions has always been a crucial issue that urgently needs to be addressed. Multiple medical institutions or research institutions can use the privacy protection method for multi-party data proposed in this embodiment of the invention to collaboratively calculate and analyze at least one of the following: patient numbers, case statistics, and disease incidence rates, without having to disclose their respective patient information. This is beneficial for medical surveys and epidemiological research. Therefore, this embodiment of the invention also provides a method for sharing medical data, including the following steps: C1: Obtain the user's statistical instructions, which are one of the following for each medical institution: number of patients, case statistics, and disease incidence rate; C2: Select at least three medical institutions as participants; C3: Using any of the above-described privacy protection methods for multi-party data, the sum of random numbers is obtained as the sum of the original data, and the random number of the last participant is the opposite of the sum of the original data. Specifically, this includes the following steps: C301: Select a cyclic group based on the number of participants and generate several different sequences of group elements; C302: Each participant generates a random number and several blind factors; C303: Add the random number to the blind factor, and pass the sums in order of the group element sequence to obtain the first accumulated result. C304: The blind factor is accumulated and passed in the order of different group element sequences to obtain the second accumulation result; C305: Subtract the second sum from the first sum to obtain the sum of random numbers; C306: Let the last participant's random number be the opposite of the sum of the above random numbers, so that the sum of the random numbers of all participants is 0; C4. Each participant standardizes the original data, and each participant adds a random number to the standardized data. The sum of the standardized data is passed in the order of the group element sequence to obtain the sum of the standardized data, which corresponds to one of the following for the user to be tested: the number of patients, case statistics, and disease incidence rate. This achieves the calculation of at least one of the following: the number of patients, case statistics, and disease incidence rate using the above-mentioned privacy protection method for multi-party data.

[0051] This invention also proposes an apparatus including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the privacy protection method for multi-party data as described in any of the above embodiments.

[0052] Figure 4This is a block diagram of a multi-party secure summation system provided in an embodiment of the present invention. The system includes a coordinating node and multiple participating node nodes. The coordinating node is responsible for selecting the cyclic group, determining the order of group elements generated by each generator, and the correspondence between each participating node and the group elements. Each participating node has the original data that needs to be kept confidential and has the ability to generate random numbers, blind factors, and perform addition operations. Different nodes can send and receive data in a certain order.

[0053] Example 1: Step S1, based on the number of participants N Choose an appropriate addition cycle group ,in The identity element is 0. The number of group elements is greater than or equal to the number of participants. Each participant corresponds to a group element, and the group elements can have different arrangements, corresponding to different orders in which the participants transmit data. For example, if n=5 and the generators are 1, 2, 3, 4, the corresponding generated group element sequences are {1,2,3,4,0}, {2,4,1,3,0}, {3,1,4,2,0}, and {4,3,2,1,0}.

[0054] Step S2, each participant generates two sets of data sequences locally, such as Figure 2 As shown, one set is random numbers. One group is used to hide the participants' original data; the other group is the blinding factor. This is used to hide its random numbers.

[0055]

[0056] Different numbers of blinding factors can be generated based on different security level requirements. The number of blinding factors generated for each participant satisfies The more items there are, the higher the security level. You can choose according to different application scenarios and needs to achieve a flexible security strategy.

[0057] Step S3, transmitting data, can also be divided into the following specific contents: S31: Each participant adds its own random number to the blind factor and then accumulates and passes the result in the order of the group elements regenerated by different generators. The first accumulation result is obtained after adding all random numbers to the blind factor. S32: Each participant will pass on their blind factors in order of the group elements regenerated by different generators, and the second sum will be obtained after adding all blind factors.

[0058] Table 1 below provides an example of generating different sequences of group elements using a generator. This example includes five participants, each corresponding to a group element and possessing an original value X, a random number, and a blind factor. In each round of data transmission, a generator is selected to generate a sequence of group elements, thus producing different arrangements of the group elements, corresponding to different orders in which the participants transmit data. After the first round of transmission, the participants... The results held are In subsequent rounds, different blinding factors are passed on in sequence, and the final participants... The results held are In the final round, the original value and the sum of the random number are passed in sequence, and the final participants... The results held are .

[0059] Table 1

[0060] Step S4: Calculate the sum of random numbers, and refer to Table 1. The participating parties... calculate The result is the sum of the random numbers from all other participants. .

[0061] Step S5, let random numbers If the result is -15 in Table 1, then the random numbers for all participants can satisfy the following condition. And neither party knows the random numbers of the other participants.

[0062] Step S6: Calculate the sum of the original data for each participant. Use your original data With random numbers Add them together and pass the result to the next participant in sequence.

[0063] Combination Figure 3 As shown, The results will be received eventually. ,because Then we can deduce The result is that the original values ​​of other participants are unknown during the process.

[0064] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, several equivalent substitutions or obvious modifications can be made without departing from the concept of the present invention, and all such modifications, achieving the same performance or purpose, should be considered within the scope of protection of the present invention.

Claims

1. A privacy protection method for multi-party data, characterized by, Includes the following steps: S1: Select a cyclic group based on the number of participants to generate a sequence of group elements in different orders; S2: Each participant generates a random number and a blinding factor. The random number is used to hide the participant's original data, and the blinding factor is used to hide the random number. S3: The sum of the random number and the blind factor is accumulated and passed in different orders according to the group element sequence to obtain a first accumulation result, and the blind factor is accumulated and passed in different orders according to the group element sequence to obtain a second accumulation result; S4: Subtract the second accumulation result from the first accumulation result to obtain the sum of the random numbers; S5: Take the opposite of the sum of the random numbers as the random number for the last participant; S6: The random number is added to the original data and transmitted to obtain the sum of the original data of the participants, thereby protecting the privacy of the original data; An additive cyclic group is selected based on the number of participants, wherein the number of group elements in the additive cyclic group is greater than or equal to the number of participants, and each participant corresponds to one group element; Regenerate the group element sequence based on the different generators of the cyclic group; Step S3 includes the following steps: S31: Each participant adds its own random number to the blind factor, and then accumulates and passes the result in the order of the group elements regenerated by different generators. After adding all the random numbers to the blind factor, the first accumulation result is obtained. S32: Each participant will pass on its blind factors in sequence according to the order of the group elements regenerated by different generators, and the second accumulation result is obtained after adding all the blind factors.

2. The privacy protection method for multi-party data as described in claim 1, characterized in that, The generator of the additive cyclic group is a set S, which consists of integers less than n and relatively prime to n, where n is a prime number greater than or equal to the number of participants; if the number of participants is not a prime number, then n is a prime number greater than and closest to the number of participants; in different sequences of group elements, the next adjacent group element of the same group element is different; the order of the group elements regenerated by different generators is different. The last group element of the sequence of group elements regenerated by different generators is the same.

3. The privacy protection method for multi-party data as described in claim 1, characterized in that, The last group element in the group element sequence corresponds to a specific participant, who calculates the sum of the random numbers.

4. The privacy protection method for multi-party data as described in claim 1, characterized in that, Each participant adds its original data to the random number to obtain ciphertext data, and broadcasts the ciphertext data to all participants. Each participant adds the ciphertext data received from other participants to its own ciphertext data to obtain the sum of the original data, so that each participant can obtain the sum of the original data.

5. The privacy protection method for multi-party data as described in claim 1, characterized in that, Each participant adds its original data to the random number to obtain ciphertext data, and then passes the ciphertext data to the next participant in a sequential transmission mode, so that only one participant receives the sum of the original data of the participants.

6. A joint credit investigation method, characterized in that, Includes the following steps: A1: Obtain the credit information request of the user to be tested; A2: Select at least three parties from the following sources as participating parties: banks, tax authorities, public security authorities, and the workplace of the user to be tested; A3: The sum of random numbers is obtained using the privacy protection method for multi-party data as described in any one of claims 1 to 5, and the random number of the last participating party is the opposite of the sum of the random numbers; A4: Each participant will standardize the raw data, and the standardized data will correspond to the credit score of the user to be investigated. A5: Each participant adds the random number to the standardized data, accumulates and transmits the results to obtain the standardized total credit score of the user to be investigated, thereby obtaining the joint credit information of the user to be investigated.

7. A method for detecting abnormal financial behavior, characterized in that, Includes the following steps: B1: Obtain instructions for checking abnormal financial behavior of the user to be detected; B2: Select at least three financial institutions as participants; B3: The sum of random numbers is obtained using the privacy protection method for multi-party data as described in any one of claims 1 to 5, and the random number of the last participating party is the opposite of the sum of the random numbers; B4: Each participant will standardize the raw data, and the standardized data will correspond to the abnormal transaction amount of the user to be detected. B5: Each financial institution adds the random number to the standardized data, accumulates and transmits the sum to obtain the standardized total amount of abnormal transactions of the user to be detected, thereby detecting whether the user to be detected has potential abnormal financial behavior.

8. A method for sharing medical data, characterized in that, Includes the following steps: C1: Obtain the user's statistical instructions, which are one of the following for each medical institution: number of patients, case statistics, and disease incidence rate; C2: Select at least three medical institutions as participants; C3: The sum of random numbers is obtained using the privacy protection method for multi-party data as described in any one of claims 1 to 5, and the random number of the last participating party is the opposite of the sum of the random numbers; C4: Each participant standardizes the original data, and each participant adds a random number to the standardized data. After accumulating and transmitting, they obtain at least one of the following standardized data for each medical institution: number of patients, case statistics, and disease incidence rate.

9. A device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the privacy protection method for multi-party data as described in any one of claims 1 to 5.

Images