Interrelated Notional Data Generation for Privacy-Safe Software Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies struggle to create complex webs of interrelated datasets within dataframes, failing to capture domain-specific randomness and interrelated properties present in real-world data, which poses challenges in generating datasets for testing software applications due to privacy concerns and security risks.
Innovation Solution
A notional data generation system that defines functional relationships between object types, adds system columns to dataframes, and fills relationship data to generate interrelated notional data, ensuring reproducibility and reduced bias.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If actual real-world data is used for testing software applications, then the realism and authenticity of test data is improved, but privacy concerns and data integrity risks worsen
Solution Approach 1:
The patent creates synthetic copies of real-world data that replicate the statistical properties, relationships, and patterns of actual data without containing sensitive information. The system generates notional data objects that mirror the structure and interrelationships of real data, providing realistic test data while eliminating privacy and security risks associated with using actual real-world data.
2Object-affected harmful factors
If simple random data is generated, then security risks are reduced, but the ability to capture domain-specific randomness and interrelated properties worsens
Solution Approach 1:
The system transforms simple random data into domain-specific realistic data by applying parameter changes that enforce functional relationships and statistical properties. The generation process incorporates domain knowledge through parameters that control data distributions, relationships between data objects, and realistic patterns, thereby maintaining security while capturing the complexity of real-world data.
Solution Approach 2:
The patent implements feedback mechanisms where generated data is continuously validated against defined functional relationships and statistical constraints. The system adjusts generation parameters based on feedback from relationship validation, ensuring that synthesized data maintains proper interrelated properties and domain-specific characteristics while remaining synthetic and secure.
3Reliability
If complex interrelated datasets are generated to capture real-world properties, then the reliability and usefulness of test data is improved, but the complexity of the data generation system worsens
Solution Approach 1:
The patent segments the data generation process into modular components: base dataframe creation, functional relationship definition, notional data object generation, and relationship population. Each component handles a specific aspect of data synthesis, making the overall complex system manageable through clear separation of concerns and reusable modules that can be independently configured and validated.
Data Source
AI summary
Disclosed herein are systems and methods for generating notional data. The method includes: receiving seed data of one or more object types in a base dataframe; defining one or more functional relationships associated with the one or more object types, at least one functional relationship of the one or more functional relationships specifying a change to seed data of one object type of the one or more object types; generating data of the one or more object types based at least in part on the seed data in the base dataframe and the one or more functional relationships; and generating the notional data based at least in part on the generated data of the one or more object types.