Machine Learning Data Collection with Approval Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Machine learning data is often inadvertently disclosed to unspecified individuals without prior approval from the data owner, leading to potential misuse and loss of control over sensitive information.
Innovation Solution
A data collection apparatus and method that requests approval from the data owner before using machine learning data, determining transmission to a database or temporary database based on similarity with previously transmitted data and approval status, ensuring that data is only shared with authorized parties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If machine learning data is automatically collected and transmitted to databases without prior approval, then data collection efficiency is improved, but data security and authorization control deteriorate
Solution Approach 1:
The system performs preliminary actions by storing approval information in advance in the information processing device. Before transmitting machine learning data, the system checks whether approval has been previously obtained. This allows the system to maintain security protocols while improving efficiency, as approved data can be transmitted without repeated approval requests, yet unauthorized data transmission is prevented through the approval verification mechanism.
2Reliability
If approval requests are sent for every machine learning data transmission, then data authorization control is improved, but processing time increases
Solution Approach 1:
The system performs preliminary actions by storing approval information in advance in the information processing device. Before transmitting machine learning data, the system checks whether approval has been previously obtained. This allows the system to maintain security protocols while improving efficiency, as approved data can be transmitted without repeated approval requests, yet unauthorized data transmission is prevented through the approval verification mechanism.
Solution Approach 2:
The system uses the previously obtained approval information as a copy or reference to verify authorization status. Instead of requiring new approval for each transmission, the system references the stored approval information, significantly reducing processing time while maintaining authorization control. The approval information acts as a reusable credential that can be verified multiple times without additional overhead.
3Adaptability or versatility
If machine learning data is transmitted to multiple databases, then data utility is improved, but risk of unauthorized disclosure increases
Solution Approach 1:
The system applies preliminary anti-action by verifying approval status before transmitting machine learning data to any database. The information processing device checks whether approval information exists that permits transmission to the target database. This preventive measure counteracts the potential harmful effect of unauthorized disclosure before it can occur, allowing the system to transmit data to multiple databases when approved while blocking transmissions when not authorized.
Data Source
AI summary
A data collection method for machine learning is proposed. The method may include accessing machine learning data stored in a device, and requesting an approval for a use of the machine learning data from an entity which has generated the machine learning data. The method may also include determining whether to transmit the machine learning data to a database for learning or a temporary database based on a result of the approval and a similarity between the machine learning data and previously-transmitted-machine learning data. The previously-transmitted-machine learning data may include a first machine learning data which has previously transmitted to the database for learning and a second machine learning data which has previously transmitted to the temporary database.