Mock Data Generation Preserving Statistical Properties
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Software providers face challenges in diagnosing and resolving technical issues encountered by users when managing or processing realistic data, as standard test data may not accurately represent user data, and accessing user data is often infeasible due to volume, confidentiality, or privacy concerns.
Innovation Solution
A computer-implemented method that generates realistic mock data by receiving statistical artefacts from an original database table, reconstructing a database table with consistent distributions and meta-distributions, and preserving foreign key relationships, allowing for the creation of a reconstructed database table that simulates the original data without exposing sensitive information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If standard test data is used for software verification, then the verification process is simple and efficient, but the test data may not be representative of realistic user data
Solution Approach 1:
The patent creates mock database tables that copy the statistical properties and distributions of original user data without copying the actual data itself. This allows realistic testing while maintaining data privacy and security, resolving the contradiction between using representative data and maintaining verification efficiency.
2Measurement precision
If user data is accessed for problem diagnosis, then accurate technical issues can be identified, but data transfer and privacy concerns arise
Solution Approach 1:
The patent introduces mock data as an intermediary that preserves the statistical characteristics and relationships of original user data while removing sensitive information. This intermediary allows accurate problem diagnosis without direct access to private user data, eliminating privacy risks while maintaining diagnostic accuracy.
Solution Approach 2:
The system creates statistical copies of user data that capture the essential patterns and distributions needed for diagnosis without containing actual sensitive information. This copying approach enables accurate technical issue identification while protecting user privacy.
3Measurement precision
If large volume user data is transferred for analysis, then comprehensive problem understanding is achieved, but data transfer becomes infeasible
Solution Approach 1:
The patent extracts only the essential statistical properties and distributions from large volumes of original user data to create compact mock data representations. This extraction process retains the analytical value needed for comprehensive problem understanding while reducing data volume to feasible transfer sizes.
Data Source
AI summary
A variety of scenarios for generating mock data based on original data are described. Characteristics of the original data can be reflected in the mock data without having access to the actual original data. Instead, statistical artifacts representing the original data can be used to instantiate reconstructed database tables. In relational database scenarios, foreign key relationships can be preserved even if the actual values are not known. The reconstructed database tables can be used as mock data in a variety of circumstances, including testing, and is especially useful to recreate characteristics of the original data when the original data cannot be made available for size, confidentiality, or privacy reasons.