Crowdsourcing platform statistical data on-demand sharing method based on blockchain and differential privacy

Abstract

The application discloses a kind of based on crowd sourcing platform statistical data on-demand sharing method of blockchain and differential privacy, differential privacy technology is added with noise design, can protect the privacy of user data, avoid data to be attacked after statistical leakage and be maliciously used.Meanwhile, the design of noise adding party and data holder dual role disperses the right of design, weakens the right concentration of supervision;Adopt blockchain technology, homomorphic encryption and zero-knowledge proof technology, ensure the integrity and non-tamperability of data, and minimize the access set of data, only allow specified regulatory party to hold the final key, so as to enhance the security and credibility of data, guarantee the availability of data;Adopt smart contract technology, can realize the automation and standardization of data, so as to meet the data compliance requirements;Key list scheme is used to classify keys, and the encrypted statistical data is identified by keyword, so as to achieve repeated use.

Description

Technical Field

[0001] This invention relates to the field of data sharing technology, and in particular to a method for on-demand sharing of statistical data from a crowdsourcing platform based on blockchain and differential privacy. Background Technology

[0002] Today, data analytics companies have a huge demand for customized data for specific industries, but the amount of such data generated has always been relatively small. The limited amount of data available is often held by leading companies in the industry and cannot be directly obtained. Therefore, using crowdsourcing platforms to conduct industry information research has become a mainstream method for collecting customized industry data.

[0003] Crowdsourcing refers to the practice of a company or organization outsourcing tasks that were previously performed by employees to a large number of non-specific (and often large) groups of volunteers on a voluntary basis. Crowdsourced tasks are usually undertaken by individuals, but if the task involves collaboration among multiple people, it may also take the form of individual production relying on open source.

[0004] In their review, Feng Jianhong et al. proposed a specific process for crowdsourcing. The main participants in crowdsourcing include task requesters and task completers. They are connected through tasks. When a task requester intends to use crowdsourcing to complete their task, they need to follow these steps: (1) Design the task; (2) Post the task on the crowdsourcing platform and wait for answers; (3) Reject or accept the worker's answers; (4) Organize the results based on the worker's answers and complete their task. The main steps for workers using crowdsourcing include: ① Finding tasks of interest; ② Accepting tasks; ③ Answering tasks; ④ Submitting answers.

[0005] In their review, Feng Jianhong et al. summarized some problems existing in the current crowdsourcing scenario: the validity of the data obtained through crowdsourcing cannot be determined, which affects the quality of the task; task completers cannot be well matched with the required tasks, and the task recommendation mechanism needs to be improved based on the preferences of task completers; there are currently no good solutions for data security and user privacy security of crowdsourcing platforms, and independent privacy of each user cannot be guaranteed for micro-complex tasks.

[0006] As seen in the above research, one solution is to use distributed storage technology to build a partially decentralized consortium blockchain, and then implement the crowdsourcing process based on this blockchain. Leveraging the immutability of blockchain, task postings and completion data are recorded; based on blockchain's traceability, the source of invalid data can be audited; and based on the accessibility of the consortium blockchain, user permission management is implemented, improving the system's reliability and security. Domestic and international research has proposed many specific solutions based on platforms such as Ethereum and Hyperledger blockchain to address the issues of data storage and privacy protection in crowdsourcing scenarios.

[0007] Li Ming et al. proposed CrowdBC, a decentralized crowdsourcing framework based on a public blockchain. In this framework, a requester's task can be solved by a group of workers without relying on any trusted third-party institution. The authors built a prototype on Ethereum using real-world datasets, demonstrating its feasibility and requiring very low transaction fees. Zhu Saide et al. proposed zkCrowd, an innovative hybrid blockchain crowdsourcing platform. zkCrowd integrates a hybrid blockchain structure and improves the consensus protocol and blockchain architecture, utilizing a dual-consensus protocol and a dual-chain architecture to ensure communication security and transaction verification reliability. Xu Xiaolong et al. proposed BPCM, a blockchain-driven crowdsourcing scheme that considers privacy protection in mobile environments. This scheme uses density-based noisy application space clustering and improved dynamic programming to cluster requesters and generate service strategies. Furthermore, it employs simple additive weighting and multi-criteria decision-making methods to optimize between maximizing service time, increasing profits, and reducing energy consumption.

[0008] However, the aforementioned prior art has the following drawbacks:

[0009] 1. User crowdsourced task data typically traverses multiple systems within the platform, resulting in lengthy and inefficient processes. Data sharing often fails to consider the legal rights of users as individual data subjects, and data holders lack adequate protection measures to ensure data security and reliability. Furthermore, the lack of oversight in compliant data sharing between organizations can lead to a proliferation of spam data. Therefore, ensuring the secure and controllable sharing of user crowdsourced task data is a crucial issue that needs to be addressed.

[0010] 2. When crowdsourcing platforms adopt conventional blockchain solutions, directly adding regulators can easily lead to an over-concentration of regulators' power. The encryption of blocks using ordinary encryption methods can also prevent the effective use of some past data. Attacks on the statistical data released after the data is collected can also cause data leakage. Summary of the Invention

[0011] This invention addresses the inherent regulatory gaps and data leakage issues of blockchain by proposing a method for on-demand sharing of statistical data in a crowdsourcing platform based on blockchain and differential privacy. In this scheme, to ensure the reuse of previously encrypted data, a key classification design is employed; to guarantee data availability, zero-knowledge proof technology and regulatory nodes are used to jointly constrain the data; to prevent attacks on statistical data, differential privacy protection measures are used to add noise; and to reduce the concentration of regulatory power, a dual-role design of noise adder and data holder is adopted to distribute power.

[0012] To achieve the above objectives, the present invention provides the following technical solution:

[0013] This invention provides a method for on-demand sharing of statistical data from a crowdsourcing platform based on blockchain and differential privacy, comprising the following sharing process:

[0014] Task data submission process: Users of the crowdsourcing platform submit data by calling the smart contract interface. Before data submission, the data is encrypted using a homomorphic encryption module. The homomorphic encryption module uses a homomorphic encryption algorithm to first encrypt the data using the data holder's key and then encrypt it using the privacy protector's key. After the data is encrypted, it is submitted to the smart contract and published on the blockchain.

[0015] Data aggregation and differential privacy protection process: The privacy protector extracts the encrypted data submitted by the user through the smart contract interface. Before using the data, the privacy protector decrypts the data using its combined key. The zero-knowledge proof module is called on the decrypted data to ensure the availability of the data. After collecting the data, the data aggregation and differential privacy protection module is called to perform homomorphic aggregation on the data, add differential privacy noise, and publish the data to the blockchain.

[0016] Data on-demand distribution process: When a data holder issues a key to a data task provider, it records the key-value pair between the key and the type of task the user is undertaking, and records it into a corresponding key list; when a third-party data requester requests data for the corresponding task, the data holder uses the keywords to collect statistical data, extract it, and deliver it to the third-party data requester.

[0017] Furthermore, before data submission, system initialization is performed. The privacy protector and data holder nodes first use the Paillier homomorphic encryption algorithm to generate a public-private key pair for the homomorphic encryption algorithm. This process is repeated for the privacy protector node to generate its own public-private key pair (pk). a ,sk a This generates a public-private key pair (pk) for a specific category for the data holder. b ,sk b The data holder also records the public / private key pair and category name in a local list; do not PK the public key.a with PK b Broadcast to existing task submitter nodes in the network, and also notify PK when a new task submitter node joins. a with PK b .

[0018] Furthermore, the method for generating the public-private key pair for the homomorphic encryption algorithm is as follows:

[0019] First, randomly select large prime numbers p and q with similar lengths, and satisfy gcd(pq, (p-1)(q-1))=1. Calculate N=pq and λ=lcm(p-1, q-1), where gcd(`) is used to calculate the greatest common factor and lcm(`) is used to calculate the least common multiple.

[0020] Then, randomly select And satisfy gcd(L(g λ mod N 2 ), N)=1, let μ=(L(g) λ modN 2 )) - 1 modN, function Let (N, g) be the node public key pair and (λ, μ) be the node private key pair.

[0021] Furthermore, during the task data submission process, user data is first proven by adding a local zero-knowledge proof module, and then homomorphic encryption is performed using the Paillier algorithm.

[0022] Furthermore, the validity of data encrypted using the Paillier algorithm is proven by combining the Pedersen, Fujisaki, and Bulletproof algorithms. The specific method is as follows:

[0023] First, define user u i The source data is v i Retrieve the Paillier algorithm public key (N, g) from the data holder; retrieve The generator g2 of the q-order subgroup and the random group element h2 in the group constitute the parameters (g2, h2, N) of the Fujisaki algorithm; take The generator g3 of the q-order subgroup and the random group element h3 in the group constitute the Pedersen algorithm parameters (g3, h3, p, q); defined as:

[0024]

[0025] Then, using the length of the range of data to be encrypted as parameter l and a random security parameter k, generate... definition:

[0026]

[0027] Definition e=H(a, b, c, x, y, z), f=d+ev,

[0028] Finally, Enc b (message)=(a, b, c, x, y, z, f, r xe r ye r ze After packaging, it is encrypted using the public key of the privacy protector to generate Enc. a (Enc b The message is then transmitted to the smart contract.

[0029] Furthermore, during the data aggregation and differential privacy protection process, the privacy protector performs aggregation and differential privacy noise addition on the data, following these steps:

[0030] Step 1: The privacy protector receives the encrypted public key Enc a (Enc b After (message)), first use your own key sk a Decrypt the data to obtain

[0031] Enc b (message)=(a, b, c, x, y, z, f, r xe r ye r ze );

[0032] Step 2: Define 0 ≤ f ≤ 2 l+2k+1 Using formula (3), the range of the data is verified:

[0033]

[0034] If the equation holds true, the verification passes, and user u is obtained. i encrypted data i ;

[0035] Step 3: After collecting a number of verified data, the privacy protection party uses the homomorphic addition property of the Paillier algorithm to aggregate the data to obtain aggregated data; then, using formula (4), noise is added to the aggregated data to obtain aggregated noise data.

[0036]

[0037] Where γ represents random noise sampling on the aggregated dataset.

[0038] Step 4: The privacy protector will aggregate the noisy data. Publish it to the blockchain and simultaneously call back the smart contract.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] The method of this invention, by employing differential privacy technology, smart contract technology in blockchain, zero-knowledge proof technology, and homomorphic encryption technology, enables secure, standardized, reliable, traceable, and accessible sharing of statistical data on crowdsourcing platforms, thereby promoting more application scenarios and data value. The main contributions of this invention include:

[0041] 1. Privacy Protection: The design employs differential privacy technology to add noise, which protects user data privacy and prevents data leakage and malicious use after statistical attacks. Simultaneously, the design uses a dual role system—the noise adder and the data holder—to distribute power and reduce the concentration of regulatory authority.

[0042] 2. Data Trustworthiness: By employing blockchain technology, homomorphic encryption, and zero-knowledge proof technology, the integrity and immutability of data are ensured, and the access set of data is minimized, allowing only designated regulatory parties to hold the final key, thereby enhancing data security and trustworthiness and guaranteeing data availability.

[0043] 3. Compliance: By adopting smart contract technology, data can be automated and standardized, thereby meeting data compliance requirements.

[0044] 4. Reusability: The key list scheme is used to classify the keys and the encrypted statistical data is identified by keywords, so as to achieve reuse.

[0045] In summary, the cryptographic technology and smart contract technology in blockchain used in this invention can solve the privacy and security issues encountered by crowdsourcing platforms in the process of data sharing, improve the security and credibility of data, and promote more application scenarios and data value. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0047] Figure 1 This is a framework diagram of a crowdsourcing platform statistical data sharing method based on blockchain and differential privacy provided in an embodiment of the present invention. Detailed Implementation

[0048] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] This invention proposes a method for on-demand sharing of statistical data in a crowdsourcing platform based on blockchain and differential privacy, such as... Figure 1 As shown, four roles are involved: the task submitter, the third-party data requester, the data holder, and the privacy protector. The third-party data requester includes the task publisher, legitimate organizations, and individuals who comply with external sharing access policies.

[0050] The on-demand sharing method for crowdsourcing platform statistics includes the following sharing process:

[0051] Task Data Submission: Users of the crowdsourcing platform submit data by calling the smart contract interface. Before submission, the data needs to be encrypted using a homomorphic encryption module. This module uses a homomorphic encryption algorithm combined with a key based on a combination of the data holder's and the privacy protector's keys. The data is first encrypted using the data holder's key, and then encrypted again using the privacy protector's key. After encryption, the data is submitted to the smart contract and published to the blockchain.

[0052] Data aggregation and differential privacy protection: The privacy protector extracts encrypted data submitted by users through a smart contract interface. Before using the data, it needs to be decrypted using the privacy protector's combined key. The zero-knowledge proof module is then called on the decrypted data to ensure its availability. After collecting a certain amount of data, the data aggregation and differential privacy protection module is called to perform homomorphic aggregation on the data, add differential privacy noise, and then publish the data to the blockchain.

[0053] On-demand data distribution: When a data holder issues a key to a data task provider, they record the key-value pair between the key and the type of task the user is undertaking, creating a corresponding key list. When a third-party data requester requests data for a specific task, the data holder can extract statistical data based on the keywords and deliver it to the third-party data requester.

[0054] The sharing process is described in detail below. The symbols used in the scheme are explained in the following table:

[0055] Table 1 Symbol Representation and Description

[0056]

[0057] (1) System initialization algorithm:

[0058] During system initialization, public-private key pairs for homomorphic encryption algorithms need to be generated on the privacy protection and data holder nodes in the system. This scheme uses the Paillier homomorphic encryption algorithm for generation. First, randomly select large prime numbers p and q with similar lengths, satisfying gcd(pq, (p-1)(q-1))=1, and calculate N=pq, λ=lcm(p-1, q-1), where gcd(`) is used to calculate the greatest common divisor and lcm(`) is used to calculate the least common multiple.

[0059] Random selection And satisfy gc d(L(g) λ mod N 2 ), N)=1, let μ=(L(g) λ modN 2 )) -1 modN, function Let (N, g) be the node public key pair and (λ, μ) be the node private key pair.

[0060] The system initialization algorithm is applied to both the privacy protector and data holder nodes to generate a public-private key pair (pk) for the privacy protector. a ,sk a This generates a public-private key pair (pk) for a specific category for the data holder. b ,sk b The data holder also records the public / private key pair and the category name in a local list. The public key is then PK'd. a with PK b Broadcast to existing task submitter nodes in the network, and also notify PK when a new task submitter node joins. a with PK b .

[0061] (2) Task data submission algorithm:

[0062] User data is first validated using a local zero-knowledge proof module, and then homomorphically encrypted using the Paillier algorithm. Since a proof needs to be generated for the encrypted data so that the privacy protector can prove its validity without accessing the original data, this solution combines Pedersen commitment, Fujisaki commitment, and Bulletproof algorithms to prove the validity of the data encrypted using the Paillier algorithm.

[0063] First, define user u i The source data is v i Retrieve the Paillier algorithm public key (N, g) from the data holder. The generator g2 of the q-order subgroup and the random group element h2 in the group constitute the parameters (g2, h2, N) of the Fujisaki algorithm. The generator g3 of the q-order subgroup and the random group element h3 in the group constitute the Pedersen algorithm parameters (g3, h3, p, q). Defined by formula (1):

[0064]

[0065] Then, using the length of the range of data to be encrypted as parameter l, and a random security parameter k, generate... Defined by formula (2):

[0066]

[0067] Define e = H(a, b, c, x, y, z). Enc b (message)=(a, b, c, x, y, z, f, r xe r ye r ze After packaging, it is encrypted using the public key of the privacy protector to generate Enc. a (Enc b The message is then transmitted to the smart contract.

[0068] (3) Data aggregation and difference algorithms:

[0069] The privacy protector aggregates and adds differential privacy noise to the data, following these steps:

[0070] Step 1: The privacy protection party receives Enc a (Enc b After (message)), first use your own key sk a Decrypt the data to obtain Enc b (message)=(a, b, c, x, y, z, f, r xe r ye r ze ).

[0071] Step 2: Define 0 ≤ f ≤ 2 l+2k+1 Using formula (3), the range of the data is verified:

[0072]

[0073] If the equation holds true, the verification passes, and user u is obtained. i encrypted data i .

[0074] Step 3: After collecting a number of verified data points, the privacy protector uses the homomorphic addition property of the Paillier algorithm to aggregate the data, obtaining aggregated data. Then, using formula 4-4, noise is added to the aggregated data to obtain aggregated noisy data.

[0075]

[0076] Where γ represents random noise sampling on the aggregated dataset.

[0077] Step 4: The privacy protector will aggregate the noisy data. Publish it to the blockchain and simultaneously call back the smart contract.

[0078] (4) Data extraction algorithm:

[0079] Data holders categorize the aggregated noisy data stored on the blockchain based on the public-private key pairs recorded in the list. When a third-party data requester submits a data request, the corresponding data is decrypted and published to the third-party data requester.

[0080] The advantages of this invention are as follows:

[0081] 1. Data Confidentiality: In traditional crowdsourcing platforms, the main nodes through which data flows are centralized platform nodes. Data is generally shared in plaintext at these nodes, and subsequent data flow is unsupervised, making it highly susceptible to leakage when the centralized node is attacked. Blockchain platforms, designed to address this centralization issue, share data in plaintext, still lacking confidentiality. This invention introduces two partially empowered regulatory bodies, encrypting data twice using a combined key. When data flows from the task submitter to the privacy protector, it is encrypted with the data holder's public key, making it unreadable from the data holder's perspective. If the privacy protector is attacked and data leakage occurs, the attacker will not obtain the plaintext user data. When data flows from the privacy protector to the data holder, it presents as statistical data. From the data holder's perspective, they cannot directly obtain the user's actual task data through statistical data. Third-party data requesters only communicate with the data holder and cannot directly join the network. Compared to traditional blockchain platforms, this invention achieves data confidentiality.

[0082] 2. Data Resistance to Statistical Attacks: In crowdsourcing platforms' on-demand sharing of statistical data, when data flows to third-party data requesters, it presents aggregated statistical data. Intuitively, it's impossible to analyze individual data from these statistical data, thus protecting personal data privacy to some extent. However, differential privacy attacks can extract user data from several statistical data sets. Therefore, this invention employs differential privacy technology. By introducing noise and controlling it within a certain privacy budget, the data retains statistical usability and eliminates the possibility of differential attacks, making the data resistant to statistical attacks during compliant queries.

[0083] 3. Process Transparency: Traditional crowdsourcing platforms primarily rely on centralized platforms for operations, resulting in opaque execution processes and an inability to ascertain whether additional operations occurred along the data's path from user to requester. This invention employs blockchain-based smart contract technology, making the smart contract code public. Any participant can view and verify the contract's functionality and execution process. This means all participants can understand the contract's specific content, avoiding distrust and disputes arising from a lack of transparency. Secondly, the smart contract's execution process is public and recorded on the blockchain. A blockchain is a decentralized distributed database where all nodes store the same data, and each block contains the hash value of the previous block, ensuring data integrity. This distributed recording method prevents interference from any centralized institution and also prevents data tampering or deletion. Finally, the smart contract's execution result is also public and recorded on the blockchain. Each node can verify the smart contract's execution result and reach consensus through a consensus algorithm. This public and consensus-based approach prevents any party from tampering with the execution result and also prevents its concealment or deletion, thus achieving process transparency.

[0084] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. However, these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for on-demand sharing of statistical data in a crowdsourcing platform based on blockchain and differential privacy, characterized in that, Includes the following shared processes: Task data submission process: Users of the crowdsourcing platform submit data by calling the smart contract interface. Before data submission, the data is encrypted using a homomorphic encryption module. The homomorphic encryption module uses a homomorphic encryption algorithm to first encrypt the data using the data holder's key and then encrypt it using the privacy protector's key. After the data is encrypted, it is submitted to the smart contract and published on the blockchain. Data aggregation and differential privacy protection process: The privacy protector extracts the encrypted data submitted by the user through the smart contract interface. Before using the data, the privacy protector decrypts the data using its combined key. The zero-knowledge proof module is called on the decrypted data to ensure the availability of the data. After collecting the data, the data aggregation and differential privacy protection module is called to perform homomorphic aggregation on the data, add differential privacy noise, and publish the data to the blockchain. Data on-demand distribution process: When a data holder issues a key to a data task provider, it records the key-value pair between the key and the type of task the user is undertaking, and records it into a corresponding key list; when a third-party data requester requests data for the corresponding task, the data holder uses the keywords to collect statistical data, extract it, and deliver it to the third-party data requester.

2. The method for on-demand sharing of statistical data in a crowdsourcing platform based on blockchain and differential privacy as described in claim 1, characterized in that, Before data submission, system initialization is performed, and the privacy protector and data holder nodes first use [the following method / function]. Homomorphic encryption algorithms generate public-private key pairs for homomorphic encryption, and the privacy protection party generates node public-private key pairs. Generate a public-private key pair corresponding to a specific category for the data holder. The data holder also records the public / private key pair and category name in a local list; do not save the public key. and Broadcast to existing task submitter nodes in the network, and notify them when new task submitter nodes join. and .

3. The method for on-demand sharing of statistical data in a crowdsourcing platform based on blockchain and differential privacy as described in claim 2, characterized in that, The method for generating public-private key pairs for homomorphic encryption algorithms is as follows: First, randomly select large prime numbers of similar length. , And satisfy ,calculate , ,in To calculate the greatest common divisor, To calculate the least common multiple; Then, randomly select And satisfy ,remember ,function ;Pick For node public key pairs, For node private key pairs.

4. The method for on-demand sharing of statistical data in a crowdsourcing platform based on blockchain and differential privacy as described in claim 1, characterized in that, During the task data submission process, user data is first proven using a local zero-knowledge proof module, and simultaneously... The algorithm performs homomorphic encryption.

5. The method for on-demand sharing of statistical data in a crowdsourcing platform based on blockchain and differential privacy as described in claim 4, characterized in that, This implementation combines the Pedersen algorithm, Fujisaki algorithm, and Bulletproof algorithm to achieve the use of The specific method for proving the validity of the encrypted data is as follows: First, define the user. The source data is Data holders Algorithm Public Key ;Pick of Generators of order subgroups Random group elements in the group The parameters that make up the Fujisaki algorithm ;Pick of Generators of order subgroups Random group elements in the group The parameters that make up the Pedersen algorithm ; definition: (1) Then, the length of the range of data to be encrypted is used as a parameter. Random security parameters ,generate ,definition: (2) definition , , ； Finally After packaging, it is encrypted again using the public key of the privacy protector to generate... Then it is transmitted to the smart contract.

6. The method for on-demand sharing of statistical data in a crowdsourcing platform based on blockchain and differential privacy as described in claim 5, characterized in that, During the data aggregation and differential privacy protection process, the privacy protector performs aggregation and differential privacy noise addition on the data, following these steps: Step 1: The privacy protector receives the encryption public key. Then, first use your own key Decrypt the data to obtain ； Step Two: Definition Using formula (3), the range of the data is verified: (3) If the equation is true, the verification passes, and the user is obtained. Encrypted data ; Step 3: After collecting a certain amount of verified data, the privacy protection party uses... The homomorphic addition property of the algorithm allows for the aggregation of data to obtain aggregated data; Then, using formula (4), noise is added to the aggregated data to obtain aggregated noise data. : (4) in Sampling of random noise on the aggregated dataset; Step 4: The privacy protector will aggregate the noisy data. Publish it to the blockchain and simultaneously call back the smart contract.

Images