Blockchain smart contract-driven decentralized ai data acquisition and annotation method

Abstract

The present invention belongs to the technical field of blockchains, and specifically relates to a blockchain smart contract-driven decentralized AI data acquisition and annotation method. The method comprises: deploying a decentralized workflow engine (DWE) to a target blockchain (S1); configuring roles and permissions (S2); a publisher creating a task according to the permissions of the roles and escrowing funds (S3); a system automatically splitting the task and allocating same (S4); workers receiving same and completing the task (S5); verifiers reviewing same and using the BFT consensus algorithm to reach a data consistency (S6); the system automatically executing payment (S7); and updating user reputation scores according to the performance (S8). The present invention uses the BFT consensus algorithm and the DWE in combination with a distributed task allocation mechanism and an AI-assisted verification tool, thereby eliminating intermediate links, reducing transaction cost, improving working efficiency and ensuring data quality, and further solving the problems of high cost, low efficiency and poor quality control of conventional methods.

Description

Blockchain smart contract driven AI data decentralized collection and labeling method TECHNICAL FIELD

[0001] The application belongs to the technical field of blockchains, and particularly relates to a blockchain smart contract driven AI data decentralized collection and labeling method. BACKGROUND

[0002] The existing AI data collection and labeling platforms generally have problems of high cost, low quality and low efficiency. These platforms usually rely on intermediary service providers to collect and label data, resulting in high service fees that reduce the final remuneration, and also limit the scale and efficiency of data collection and labeling work. The traditional financial payment system also limits the effective use of labor worldwide, increasing the complexity of payment rewards. In addition, the quality control methods of these platforms are relatively single, lack flexible and efficient review mechanisms, and cannot guarantee the quality of data collection and labeling.

[0003] From the above, a significant technical problem existing at present is the cost increase due to high intermediary fees of the existing platforms, and the data quality problem due to the lack of effective quality control mechanisms. SUMMARY

[0004] The purpose of the present application is to provide a blockchain smart contract driven AI data decentralized collection and labeling method, which aims to reduce transaction costs by eliminating intermediate links, improve work efficiency, and ensure data consistency and quality through the BFT data consensus algorithm, to solve the problems raised in the background technology.

[0005] To achieve the above object, the application adopts the following technical scheme: A blockchain smart contract driven AI data decentralized collection and labeling method, comprising the following steps: S1: deploying a decentralized workflow engine to a target blockchain, the workflow engine being used to manage the distribution and verification of tasks; S2: configuring roles and permissions according to the workflow engine deployed in S1, including but not limited to the permissions of workers, verifiers, auditors and publishers; S3: the publisher creates a task according to the role permissions configured in S2, uploads a data set to be collected and labeled, sets the collection and labeling requirements, and hosts corresponding funds through a smart contract; S4: the system automatically splits the original data set into multiple small tasks according to the task created in S3, and assigns them to appropriate workers according to the role configuration in S2; S5: the workers accept the tasks and collect and label the data according to the tasks assigned in S4; S6: the verifier audits the collection and labeling results submitted by the workers in S5, and reaches data consistency through a BFT consensus algorithm; S7: the system automatically executes the payment process through a smart contract according to the data consistency reached in S6; S8: update the user's credit score according to the performance of the workers in S5 to S7.

[0006] Preferably, the workflow engine deployment in S1 includes establishing a network layer, a workflow layer and a blockchain layer, and the layers communicate and coordinate with each other.

[0007] Preferably, the role and permission configuration in S2 includes RBAC permission management.

[0008] Preferably, when the publisher creates a task in S3, it also includes setting the task type, detailed task description, collection / labeled specification, reward and incentive mechanism parameters.

[0009] Preferably, the automatic task splitting and distribution in S4 also includes data analysis and preprocessing, intelligent task splitting, and generating a unique task ID.

[0010] Preferably, when the workers execute the task in S5, it also includes using AI-assisted collection and labeling tools.

[0011] Preferably, the verification and audit in S6 also includes cross-verification and voting by multiple verifiers to reach BFT consensus, and random spot checks.

[0012] Preferably, the payment and settlement in S7 also includes supporting cryptocurrency and fiat currency payments, and implementing payment between different blockchains through cross-chain bridging.

[0013] The technical effects and advantages of the application are as follows:

[0014] The present application eliminates intermediate links, reduces transaction costs, improves work efficiency, and ensures data consistency and quality through the BFT data consensus algorithm, by using the BFT data consensus algorithm and the decentralized workflow engine (DWE), combining a distributed task allocation mechanism and AI-assisted collection, labeling and verification tools. This method not only improves the overall efficiency of data collection and labeling, but also simplifies the payment process through the automatic payment process of the smart contract, thereby solving the problems of high cost, low efficiency and poor quality control in the data collection and labeling process in the prior art. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 is a flowchart of a blockchain smart contract driven AI data decentralized collection and labeling method according to the present application;

[0016] Figure 2 is a server-side architecture diagram according to the present application;

[0017] Figure 3 is a DWE system architecture diagram according to the present application;

[0018] Figure 4 is a verification and audit flowchart according to the present application. Embodiments of the present application

[0019] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The specific embodiments described herein are only used to explain the present application, and are not used to limit the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor are within the scope of protection of the present application.

[0020] The present application provides a blockchain smart contract driven AI data decentralized collection and labeling method, as shown in Figure 1, which includes the following steps:

[0021] S1: deploy a decentralized workflow engine (DWE) to the target blockchain, the workflow engine is used to manage the distribution and verification of tasks;

[0022] Specifically, as shown in Figure 2, the server side includes a workflow engine (DWE), which is responsible for managing and scheduling different functional components, such as user management, task management, API management, login / registration, role management and payment management. The client sends a request to the server to call these services.

[0023] In the server side, there are various sub-modules in the DWE module, which correspond to different functions respectively:

[0024] User Management: Handles user-related operations such as account information maintenance, identity authentication, etc.

[0025] Task Management: Assigns, tracks, and monitors tasks to ensure they are executed according to predetermined processes.

[0026] API Management: Provides Application Programming Interfaces (APIs) that allow external programs or services to interact with the system.

[0027] Login / Registration: Handles user login and registration processes, ensuring security and accessibility.

[0028] Role Management: Defines and manages different types of user roles for easy permission control and access policies.

[0029] Payment Management: Handles payment-related transactions, including payments, refunds, etc.

[0030] In the interaction between the server and the client, the client can invoke these services through requests. The smart contracts on the server are running on multiple chains, which means they can execute in different blockchain networks, providing greater flexibility and interoperability.

[0031] Further, as shown in Figure 3, the workflow engine deployment includes establishing a network layer, a workflow layer, and a blockchain layer, and each layer communicates and coordinates with each other.

[0032] Network Layer: Composed of multiple nodes and connections, primarily handling communication between nodes. Each peer connection is responsible for communication with other nodes and manages message transmission and reception through outbound and inbound queue processors. PeerConnection: Each node's connection is responsible for communication and data exchange, ensuring efficient transmission of network messages. OutboundServer & InboundServer: Handle the sending and receiving of information. The outbound server is responsible for sending information to other nodes, while the inbound server is responsible for receiving and processing information from external nodes. P2PNode: The P2P node handles messages from other nodes and distributes them to relevant inbound message processors. This module is also responsible for interacting with the blockchain layer.

[0033] Workflow Layer: The workflow layer consists of a workflow engine and a task list user interface (UI). The workflow engine is the core task allocation and scheduling component, while the UI interface provides users with visual display and interactive functions of task status. Worklist UI: Provides a graphical user interface where users can view the list of current tasks, task status, and interact with the workflow engine. Workflow Engine: The core component that manages and schedules tasks. The workflow engine distributes tasks to appropriate workers based on system-set rules and task requirements, and tracks the progress of task execution in real time.

[0034] Blockchain Layer: Mainly responsible for on-chain recording of tasks and transactions, and interaction with smart contracts. The blockchain layer ensures that important steps in the task execution process are recorded on the blockchain, ensuring transparency and immutability. Transaction Service: Manages task-related transactions to ensure that each task execution is recorded on the chain, ensuring transparency. Smart Contract Service: Automates task rules through smart contracts to ensure that tasks are processed on the platform according to predetermined logic.

[0035] S2: Configure roles and permissions based on the workflow engine deployed in S1, including but not limited to worker, verifier, auditor, and publisher permissions; further, role and permission configuration includes RBAC-based permission management to ensure the security and effectiveness of task allocation and execution.

[0036] S3: The publisher creates tasks according to the role permissions configured in S2, uploads the data set to be collected and labeled, sets the collection and labeling requirements (such as collection and labeling types), and hosts the corresponding funds through smart contracts; further, in this process, the publisher needs to set the following specific parameters:

[0037] Task Type Selection: Clearly specify whether the task is data collection, data labeling, or both, so that the system can perform corresponding processing.

[0038] Detailed Task Description: Provide specific requirements for the task, including data format, quality standards, quantity requirements, and completion deadline.

[0039] Collection / Labelling Specification: Upload or link to detailed operation guidelines to guide workers to correctly perform tasks.

[0040] Reward and Incentive Mechanism: Set the basic remuneration for each sub-task, and can set additional reward conditions such as early completion or high-quality delivery.

[0041] Fund escrow submission: According to the task size and remuneration standard, the corresponding amount of cryptocurrency is escrowed into the system through the smart contract, ensuring the safety and transparency of the funds.

[0042] S4: The system automatically splits the original data set into multiple small tasks according to the tasks created in S3, and assigns them to appropriate workers according to the role configuration in S2. Further, the automatic task splitting and assignment also includes data analysis and preprocessing, intelligent task splitting, and generating a unique task ID for tracking and management. The specific process of splitting is as follows:

[0043] Data analysis and preprocessing: Analyze the uploaded data set to determine the smallest task unit that can be split (such as an image, an audio clip, a text segment, a video segment, etc.).

[0044] Intelligent task splitting: Generate a globally unique task ID (UUID) for each small task to facilitate subsequent tracking and management. According to the worker's skill tags, historical performance, and reputation score, use a recommendation algorithm (collaborative filtering or machine learning model, etc. Different strategies) to select the most suitable worker.

[0045] S5: Workers accept tasks and collect, label data according to the tasks assigned in S4; further, workers performing tasks also include using AI-assisted collection and labeling tools. Specifically, workers perform tasks as follows:

[0046] Worker accepts task: Workers view and accept tasks assigned to themselves on the platform, read task requirements and collection and labeling guidelines.

[0047] Data collection and labeling process: Workers use the AI-assisted collection and labeling tools provided by the platform to collect and label data according to task requirements. After completing the collection and labeling, submit the collection and labeling results to the platform.

[0048] S6: Verifiers verify the collection and labeling results submitted by workers in S5, and reach data consistency through BFT consensus algorithm; further, verification and review also include cross-verification and voting by multiple verifiers to reach BFT consensus, and random spot checks.

[0049] Specifically, as shown in Figure 4, the verification and review process is as follows:

[0050] Verification and review of collection and labeling results: Verifiers receive the collection and labeling results submitted by workers, review the results according to the quality standards of the task requirements, and confirm the accuracy and consistency of the collection and labeling.

[0051] Data BFT consensus: multiple verifiers cross-verify and vote on the same data package on the blockchain to reach BFT consensus on a single data, and the voting and consensus results are divided into pass and reject verification.

[0052] Auditor sampling: auditors randomly sample part of the verified collection and labeling results to ensure the overall collection and labeling quality. For problems found, auditors will feedback to verifiers and workers, and adjust their reputation scores if necessary.

[0053] S7: The system automatically executes the payment process according to the data consistency reached in S6 through the smart contract; further, the payment and settlement also include supporting cryptocurrency and fiat currency payment, and realizing payment between different blockchains through cross-chain bridging.

[0054] Specifically, when the collection and labeling results pass the verification and audit, the smart contract automatically triggers the payment process. The reward is paid to the workers and verifiers according to the pre-set distribution ratio. The platform supports cryptocurrency and fiat currency payment. Through cross-chain bridging, the platform realizes payment and settlement between different blockchains, ensuring the convenience and timeliness of payment worldwide.

[0055] In another embodiment, to ensure high-quality completion of tasks, the system introduces a smart contract-based pledge guarantee and punishment mechanism. All participants need to pay a certain amount of pledge cryptocurrency in the smart contract as a guarantee for the quality of work to prevent malicious behavior. The amount of pledge can be dynamically adjusted according to the complexity of the task and the reputation score of the worker. The pledge funds are automatically managed by the smart contract to ensure the safety and non-tamperability of the funds.

[0056] In addition, platform participants need to pledge a certain amount of funds to the smart contract as a guarantee for the quality of work. If there is malicious work behavior (such as uploading irrelevant data, random labeling and verification), the pledged assets will be punished by the system.

[0057] S8: Update the reputation score of the user according to the performance of the worker in S5 to S7. Specifically, adjust the reputation score of the workers and verifiers according to their work quality. High-reputation-score users will have more task opportunities and higher rewards, and have the opportunity to become auditors. New users need to pass the labeling standard test (LST) to obtain the corresponding identity authentication, and the system issues a DID (decentralized identity) to identify the ability and qualification of the user who passes the test, and maintaining good work quality on a daily basis can also increase the level of DID.

[0058] In summary, the present application eliminates intermediate links, reduces transaction costs, improves work efficiency by using BFT data consensus algorithm and decentralized workflow engine (DWE), combining distributed task allocation mechanism and AI assisted collection, labeling verification tool, and ensures the consistency and quality of data through BFT data consensus algorithm. This method not only improves the overall efficiency of data collection and labeling, but also simplifies the payment process of rewards through the automatic payment process of smart contract, thereby solving the problems of high cost, low efficiency and poor quality control in the process of data collection and labeling in the prior art.

[0059] Finally, it should be noted that: the above only for the preferred embodiments of the present application, and not for limiting the present application, although the foregoing embodiments of the present application are described in detail, for those skilled in the art, it still can be modified, or part of the technical features of the equivalent replacement, within the spirit and principles of the present application, any modification, equivalent replacement, improvement, etc., should be included within the scope of the present application.

Claims

1. A method for AI data decentralized collection and labeling driven by a blockchain smart contract, characterized in that, Comprising the following steps: S1: Deploy a decentralized workflow engine to the target blockchain, which is used to manage the distribution and verification of tasks; S2: Configure roles and permissions according to the workflow engine deployed in S1, including but not limited to the permissions of workers, verifiers, auditors, and publishers; S3: The publisher creates a task according to the role permissions configured in S2, uploads the data set to be collected and labeled, sets the collection and labeling requirements, and hosts the corresponding funds through a smart contract; S4: The system automatically splits the original data set into multiple small tasks according to the task created in S3, and assigns them to appropriate workers according to the role configuration in S2; S5: Workers accept tasks and collect and label data according to the tasks assigned in S4; S6: The verifier audits the collection and labeling results submitted by the workers in S5, and reaches data consistency through the BFT consensus algorithm; S7: The system automatically executes the payment process through the smart contract according to the data consistency reached in S6. 2.The AI data decentralized collection and labeling method driven by a blockchain smart contract according to claim 1, wherein, The workflow engine deployment in S1 includes establishing network layer, workflow layer and blockchain layer, and each layer communicates and coordinates with each other. 3.The AI data decentralized collection and labeling method driven by a blockchain smart contract according to claim 1, wherein, The role and permission configuration in S2 includes permission management through RBAC. 4.The AI data decentralized collection and labeling method driven by a blockchain smart contract according to claim 1, wherein, In S3, the publisher also includes setting task type, detailed task description, collection / labeled specification, reward and incentive mechanism parameters when creating a task. 5.The AI data decentralized collection and labeling method driven by a blockchain smart contract according to claim 1, wherein, In S4, the automatic task splitting and distribution also includes data analysis and preprocessing, intelligent task splitting, and generating a unique task ID. 6.The AI data decentralized collection and labeling method driven by a blockchain smart contract according to claim 1, wherein, In S5, the worker also includes using AI-assisted collection and labeling tools when performing tasks.

7. The AI data decentralized collection and labeling method driven by a blockchain smart contract according to claim 1, characterized in that, In S6, verification and audit also include cross-verification and voting by multiple verifiers to reach BFT consensus, and random spot checks. 8.The AI data decentralized collection and labeling method driven by a blockchain smart contract according to claim 1, wherein, In S7, payment and settlement also include supporting cryptocurrency and fiat currency payments, and implementing payment between different blockchains through cross-chain bridging.

Images