User Identity Field Conversion Using Feature Rules
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for identifying user identity information in large datasets are inefficient and insecure, with high error rates and low accuracy due to complex data structures and varying data ranges.
Innovation Solution
A method involving the selection of positive and negative sample data to calculate features, establishing a feature rule for user identity, and scanning data tables to match and convert user identity data into third-party accounts, while keeping non-user identity data unchanged, using a terminal with a processor and non-volatile storage medium.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If fuzzy searching or matching based on traversing all registration data is used to identify user identity, then the method can handle complex data structures, but the efficiency is low and error rate is high
Solution Approach 1:
The patent pre-calculates and stores feature values (such as data type, length, format, value range) for user identity fields during data ingestion or preprocessing. When identification is needed, the system directly compares query data against these pre-established feature rules without traversing all registration data, thus resolving the contradiction between accuracy and efficiency
Solution Approach 2:
The patent transforms the identification approach from string-based fuzzy matching to parameter-based exact matching by extracting and comparing structural parameters (data type, length, format, value range). This parameter transformation enables both high accuracy through precise matching and high efficiency through direct parameter comparison without full data traversal
2Productivity
If data value range limiting is used to capture user identity, then the efficiency improves, but the correct user identity cannot be accurately captured because data range varies significantly
Solution Approach 1:
The patent creates a universal feature rule framework that can adapt to different user identity types (phone numbers, email addresses, ID cards, etc.) by defining multiple feature dimensions (data type, length, format, value range). This multi-functional rule system handles varying data ranges accurately while maintaining efficient identification across different identity types
3Measurement precision
If manual identification or overwriting of user identity data is attempted, then the accuracy may be high, but the process is impractical due to the massive size of data tables reaching hundreds of TBs
Solution Approach 1:
The patent implements an automated identification system that self-services the entire identification process. The system automatically extracts features, compares data against feature rules, and identifies user identity fields without human intervention. This automation makes the process operationally feasible for massive data tables that would be impossible to handle manually
Data Source
Figure 1A~1B
Figure 2~3
Figure 4
AI summary
A method for automatically converting electronic data is disclosed. The method comprises scanning a source data table containing data fields; determining a feature for each of the data fields of the source data table; comparing the feature for each of the data fields with a feature rule for identifying user-identity-containing data fields in the source data table; identifying a first data field of the source data table as containing user identity when the feature of the first data field matches the feature rule; identifying a second data field of the source data table as containing no user identity when the feature of the second data field fails to match the feature rule; converting the source data table by replacing data items of the first data field in the source data table identified as containing user identity with corresponding third-party user accounts, and keeping the second data field in the source data table identified as not containing user identity unaltered; and storing the converted data table in a storage medium.