Training Data Permission Control Using Verified Source Records
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The lack of effective control over the source of training data used in artificial intelligence models leads to disputes regarding ownership, which can arise during the use of these models.
Innovation Solution
A system and method to control the usage permission of training data by verifying the source through source verification information, enabling usage permission if the source meets predetermined conditions and disabling it otherwise, using blockchain and digital verification technologies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If training data is used without source verification, then productivity is improved by faster model training, but reliability deteriorates due to ownership disputes and unauthorized usage
Solution Approach 1:
The system performs source verification and permission checking before the training process begins. The processor obtains source verification information, confirms the data source, and determines usage permission in advance, ensuring that only authorized data is used for training. This preliminary verification prevents ownership disputes while maintaining training efficiency.
2Reliability
If source verification is implemented for all training data, then reliability is improved by preventing ownership disputes, but device complexity increases due to additional verification systems
Solution Approach 1:
The system introduces a verification information record as an intermediary element that stores source verification information. This record acts as a mediator between the training data and the verification process, containing pre-stored source confirmation data that simplifies the verification operation without requiring complex real-time analysis systems.
3Reliability
If usage permission is strictly controlled through verification, then reliability is improved by ensuring legal compliance, but productivity decreases due to additional verification steps
Solution Approach 1:
The system determines usage permission before the training process starts by obtaining source verification information and confirming data sources in advance. This preliminary permission check ensures legal compliance is established upfront, allowing the training process to proceed efficiently without interruptions or repeated verification steps during training.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
A method and system for controlling a usage permission of training data (201(1) to 201(n), 301(1) to 301(n)) are provided. The method includes: obtaining source verification information (302(1) to 302(n)) corresponding to first training data; confirming a source of the first training data based on the source verification information (302(1) to 302(n)); enabling a usage permission corresponding to the first training data if the source of the first training data meets a predetermined condition and training a computing model (101) based on the first training data; and disabling the usage permission corresponding to the first training data if the source of the first training data does not meet the predetermined condition.