Multi-party data cooperative computing method for privacy protection

Abstract

The invention relates to the technical field of privacy protection, and provides a multi-party data collaborative computing method for privacy protection, which comprises the following steps: S1, a data transaction platform designs a seller data demand library Q1 and a plurality of contracts C, and each contract C represents different privacy protection levels n and indicates a data transaction price; s2, each data owner selects a contract to sign with the data transaction platform by taking own utility maximization as a target and combining own privacy preference, and uploads original data to the data transaction platform; s3, the data transaction platform receives the original data provided by the data owner, randomly disturbs the original data according to the contract C selected by the data owner and a disturbance mechanism M, and uploads disturbance data; s4, the data transaction platform collects disturbance data and pays for each data owner according to the data transaction price signed in the corresponding contract C; and S5, the data transaction platform designs a buyer data demand library Q2.

Description

Technical Field

[0001] The present invention relates to the field of privacy protection technology, and in particular to a privacy-preserving multi-party data collaborative computing method. Background Art

[0002] Data trading exposes data owners to the risk of privacy leakage. Currently, differential privacy technology has been widely used to protect data privacy, allowing us to accurately quantify the extent of privacy leakage. In the data trading process using centralized differential privacy (CDP), a trusted data collector collects raw data from data owners, then executes a privacy protection scheme on the collected data, and finally sends the perturbed data to data users according to their requirements.

[0003] However, once these data privacy are maliciously spread by service providers, they will cause direct damage to the user's security in all aspects. Even if the service provider does not directly leak the user's original data, the attacker may attack the global model after centralized training, thereby stealing the data information involved in the training, resulting in the data owner's willingness to contribute to the data gradually decreasing. Summary of the invention

[0004] In view of the defects in the prior art, the purpose of the present invention is to provide a privacy-preserving multi-party data collaborative computing method that can maximize data utility while protecting data privacy.

[0005] The present invention provides a privacy-preserving multi-party data collaborative computing method, comprising the following steps:

[0006] S1. The data trading platform designs a seller data demand library Q1 and multiple contracts C. Each contract C represents a different privacy protection level n and indicates a data transaction price.

[0007] S2. Each data owner selects a contract to sign with the data trading platform with the goal of maximizing its own utility and in combination with its own privacy preference, and uploads the original data to the data trading platform;

[0008] S3. The data trading platform receives the original data provided by the data owner, and randomly perturbs the original data according to the perturbation mechanism M according to the contract C selected by the data owner, and uploads the perturbed data;

[0009] S4. The data trading platform collects the disturbance data and pays each data owner a reward according to the data trading price signed in the corresponding contract C;

[0010] S5. Data trading platform designs buyer data demand library Q2;

[0011] S6. The data trading platform performs matching in the data demand library Q2 and the data demand library Q1. If the matching is successful, the corresponding disturbance data is sent to the corresponding data user; if the matching fails, the matching continues.

[0012] Further, in step S3, the perturbation mechanism M is a function whose domain and range are both X. Given any data x∈X and a random perturbation probability p, the original data x is perturbed by the perturbation mechanism M with probability p equal to the original data x, and with probability is equal to some other value x′, x′≠x, x′∈X.

[0013] Furthermore, the random disturbance probability p>0.5.

[0014] Furthermore, the utility calculation formula of the data owner is:

[0015] u i =p i -c i e i ;

[0016] Wherein, pi is the remuneration obtained by the data owner i for trading the original data, ci is the privacy cost caused by the privacy leakage of a unit quantity, and ciei is the privacy cost of the data owner;

[0017] In order to maximize the utility of the data owner, the following accuracy definition is adopted: the accuracy of an estimated value λ′ of the true value λ is (ɑ, β) if the distance between it and the true value λ does not exceed 1-ɑ, ɑ≥0, 0≤β≤1, that is, the following inequality holds:

[0018] Pr{|λ′-λ|≥ɑ}≤1-β;

[0019] If the above inequality is satisfied, it means that the utility calculation result of the data owner falls near the true value with a high probability.

[0020] Beneficial effects: The privacy-preserving multi-party data collaborative computing method provided by the present invention can randomly perturb the original data according to the perturbation mechanism M according to different privacy protection levels. The data owner can comprehensively weigh different privacy protection levels and maximum utility and sign a suitable contract C with the data trading platform, thereby maximizing the data utility while protecting data privacy. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the process structure of the present invention. DETAILED DESCRIPTION

[0022] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0023] In the present invention, unless otherwise clearly specified and limited, the terms "connection" and "fixation" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0024] In the description of the present invention, it is necessary to understand that the terms "longitudinal", "lateral", "horizontal", "top", "bottom", "up", "down", "inside" and "outside" etc. indicating orientations or positional relationships are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0025] In addition, the terms "first", "second", etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. In the description of the present invention, the meaning of "plurality" is more than two, unless otherwise clearly and specifically defined.

[0026] like Figure 1 As shown, the present invention provides a privacy-preserving multi-party data collaborative computing method, comprising the following steps:

[0027] S1. The data trading platform designs the seller's data demand library Q1 and multiple contracts C. Each contract C represents a different privacy protection level n and indicates the data transaction price.

[0028] S2. Each data owner chooses a contract to sign with the data trading platform with the goal of maximizing his or her own utility and combining his or her own privacy preferences, and uploads the original data to the data trading platform.

[0029] S3. The data trading platform receives the original data from the data owner, and randomly perturbs the original data according to the perturbation mechanism M based on the contract C selected by the data owner, and uploads the perturbed data. In order to protect data privacy, we perturb the original data of the data owner through the perturbation mechanism M. Therefore, the data owner can independently control the degree of privacy leakage, and the data finally uploaded is the perturbed data.

[0030] S4. The data trading platform collects disturbance data and pays each data owner according to the data transaction price signed in the corresponding contract C, that is, the data owner obtains utility.

[0031] S5. The data trading platform designs the buyer’s data demand library Q2.

[0032] S6. The data trading platform performs matching in the buyer's data demand library Q2 and the seller's data demand library Q1. If the matching is successful, the corresponding disturbance data is sent to the corresponding data user; if the matching fails, the matching continues.

[0033] In the present invention, each data owner can process his own data independently, that is, the data privacy protection process is transferred from a hypothetical trusted third-party data collector to a single terminal device, without the intervention of a third party, thereby avoiding privacy attacks that may be caused by untrusted third-party data collectors. Each data owner uses the perturbation mechanism M to perturb the original data before conducting data transactions, and the privacy protection level and security of the data are better.

[0034] The present invention can randomly perturb the original data according to the perturbation mechanism M according to different privacy protection levels. The data owner can comprehensively weigh different privacy protection levels and maximum utility and sign a suitable contract C with the data trading platform, thereby maximizing the data utility under the premise of protecting data privacy.

[0035] In one embodiment, in step S3, the perturbation mechanism M is a function whose domain and range are both X. Given any data x∈X and a random perturbation probability p, the original data x is perturbed by the perturbation mechanism M with probability p equal to the original data x, and with probability is equal to some other value x′, x′≠x, x′∈X.

[0036] In one embodiment, the random perturbation probability p>0.5.

[0037] Without loss of generality, we can assume that p>0.5 and the privacy protection level This formula indicates the quantitative relationship between the random perturbation probability p and the corresponding privacy protection level n. If the data owner signs a contract C with a privacy protection level n with the data trading platform, the random perturbation probability that should be selected when applying the above random perturbation mechanism is Obviously, when p is closer to 0.5, the randomness of the submitted data is greater, the privacy protection level n is higher, but the corresponding data utility is also lower.

[0038] In one embodiment, generally speaking, the privacy cost brought to the data owner by the data transaction process is related to the privacy protection level n: the higher the privacy protection level n, the lower the possible privacy leakage and the lower the privacy cost; the lower the privacy protection level n, the higher the possible privacy leakage and the higher the privacy cost. In the present invention, the utility calculation formula of the data owner is:

[0039] u i =p i -c i e i ;

[0040] Among them, pi is the reward obtained by data owner i for trading original data, ci is the privacy cost caused by the privacy leakage of a unit number, and ciei is the privacy cost of the data owner;

[0041] In order to maximize the utility of the data owner, the following accuracy definition is adopted: the accuracy of an estimated value λ′ of the true value λ is (ɑ, β) if the distance between it and the true value λ does not exceed 1-ɑ, ɑ≥0, 0≤β≤1, that is, the following inequality holds:

[0042] Pr{|λ′-λ|≥ɑ}≤1-β;

[0043] If the above inequality is satisfied, it means that the utility calculation result of the data owner falls near the true value with a higher probability.

[0044] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention.

[0045] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A privacy-preserving multi-party data collaborative computing method, characterized by: The following steps are involved: S1. The data trading platform designs the seller's data demand library Q1 and multiple contracts C. Each contract C represents a different privacy protection level n and indicates the data transaction price; S2. Each data owner selects a contract to sign with the data trading platform with the goal of maximizing its own utility and in combination with its own privacy preference, and uploads the original data to the data trading platform; S3. The data trading platform receives the original data provided by the data owner, and randomly perturbs the original data according to the perturbation mechanism M according to the contract C selected by the data owner, and uploads the perturbed data; S4. The data trading platform collects the disturbance data and pays each data owner a reward according to the data trading price signed in the corresponding contract C; S5. Data trading platform designs buyer data demand library Q2; S6. The data trading platform performs matching in the buyer's data demand library Q2 and the seller's data demand library Q1. If the matching is successful, the corresponding disturbance data is sent to the corresponding data user; if the matching fails, the matching continues.

2. The privacy-preserving multi-party data collaborative computing method according to claim 1, characterized in that: In step S3, the perturbation mechanism M is a function whose domain and range are both X. Given any data x∈X and a random perturbation probability p, the original data x is perturbed by the perturbation mechanism M with probability p equal to the original data x, and with probability is equal to some other value x′, x′≠x, x′∈X.

3. The privacy-preserving multi-party data collaborative computing method according to claim 2, characterized in that: The random disturbance probability p>0.

5.

4. The privacy-preserving multi-party data collaborative computing method according to claim 1, characterized in that: The utility calculation formula of the data owner is: u i =p i -c i e i ; Wherein, pi is the remuneration obtained by the data owner i for trading the original data, ci is the privacy cost caused by the privacy leakage of a unit quantity, and ciei is the privacy cost of the data owner; In order to maximize the utility of the data owner, the following accuracy definition is adopted: the accuracy of an estimated value λ′ of the true value λ is (ɑ, β) if the distance between it and the true value λ does not exceed 1-ɑ, ɑ≥0, 0≤β≤1, that is, the following inequality holds: Pr{|λ′-λ|≥ɑ}≤1-β; If the above inequality is satisfied, it means that the utility calculation result of the data owner falls near the true value with a high probability.

Images