Anonymous User Information Generation for Privacy-Preserving Prediction Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data isolation and islanding effects are becoming more serious, leading to ineffective data sharing and prediction models that are less accurate due to the lack of data interaction between different subjects or applications.
Innovation Solution
A method and apparatus for generating prediction information by creating anonymous user information based on first user characteristic information and sending it to a second processing end to generate prediction information using both anonymous and second user characteristic information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data is shared between different processing ends, then prediction accuracy is improved, but data security and privacy protection are compromised
Solution Approach 1:
The patent introduces anonymous user information as an intermediary between first user characteristic information and second user characteristic information. This intermediary enables data sharing for improved prediction accuracy while preventing direct exposure of sensitive user data, thus maintaining data security. The anonymous user information acts as a bridge that allows collaborative modeling without compromising privacy.
Solution Approach 2:
The patent creates anonymous user information as a copy or representation of the original user characteristic information. This copy contains sufficient information for prediction purposes but excludes sensitive personal identifiers. By using this copied anonymous version instead of original data, the system achieves data sharing while protecting data security and user privacy.
2Reliability
If user characteristic information is shared between different subjects, then model effectiveness is improved, but data isolation and islanding effects are worsened
Solution Approach 1:
The anonymous user information serves as a mediator that enables data sharing between different processing ends, breaking down data silos and islanding effects. It allows models to be trained on diverse data sources while maintaining a standardized, privacy-preserving format that enhances both model effectiveness and data sharing capability simultaneously.
Solution Approach 2:
The patent transforms user characteristic information into anonymous user information by changing its parameters - removing sensitive identifiers while preserving predictive features. This parameter transformation enables the data to be shared across different subjects and applications, improving both model effectiveness and data interoperability without compromising security.
3Object-affected harmful factors
If anonymous user information is generated and transferred, then data security is maintained, but information loss may occur
Solution Approach 1:
The anonymous user information is designed with local quality - it contains specific information needed for prediction tasks while deliberately excluding sensitive personal identifiers. Different parts of the information have different properties: predictive features are preserved while personal identifiers are removed. This selective information retention maintains data security while minimizing information loss for the specific purpose of prediction.
Solution Approach 2:
The patent applies partial action by selectively transforming only the sensitive portions of user characteristic information while leaving the predictive features intact. Instead of anonymizing all data (which would cause excessive information loss), it selectively anonymizes only what is necessary for security, preserving the information needed for effective prediction.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided are a method and an apparatus for generating prediction information, an electronic device, and a computer readable medium. The method includes: generating, based on first user characteristic information of a target user, anonymous user information of the target user (201); sending the anonymous user information to a second processing end to enable the second processing end to generate prediction information based on the anonymous user information and second user characteristic information (202). Data interaction and sharing are realized while ensuring data privacy, thereby improving accuracy of the prediction information.