Data Anonymization via Clustering and Sharding for Privacy Compliance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cloud storage systems face challenges in securely storing personal data while adhering to territorial privacy laws, as data protection laws restrict the transfer of encrypted data across borders, and the risk of data leakage remains high.
Innovation Solution
The method involves anonymizing input data by clustering data points into groups, assigning unique identifiers, and generating translation, deviation, cluster, and reference tables, which are then stored in a way that facilitates secure and rapid data retrieval, while maintaining data sovereignty.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is encrypted and stored in cloud, then data security is improved, but data leakage risk increases due to potential key compromise or algorithm advances
Solution Approach 1:
The patent segments data into multiple shards and stores them across different geographic locations. Each shard alone is insufficient to reconstruct the original data, thereby reducing the risk that a single breach or key compromise will lead to complete data leakage. This segmentation approach maintains security while enabling cloud storage.
Solution Approach 2:
The patent introduces secret sharing mechanisms and distributed key management as intermediaries between the data and storage system. Instead of storing a single encryption key, the system uses multiple key shares distributed across different locations, requiring a threshold number of shares to reconstruct the original data. This intermediary layer prevents direct access even if some storage nodes are compromised.
2Reliability
If data is stored locally to meet territorial privacy laws, then data sovereignty is maintained, but cloud storage flexibility and load balancing are restricted
Solution Approach 1:
The patent divides data into geographic segments or shards that can be stored in different jurisdictions. By segmenting data across multiple locations while maintaining control through secret sharing, the system achieves both data sovereignty (through controlled access) and cloud storage flexibility (through distributed storage and load balancing capabilities).
Solution Approach 2:
The patent adds a new dimension to data storage by implementing hierarchical or multi-layered storage architecture. Data can be stored in cloud infrastructure while access control and sovereignty requirements are enforced through additional layers of secret sharing and key management, effectively resolving the conflict between physical location constraints and cloud flexibility.
3Reliability
If data is anonymized through clustering and multiple tables, then data security and anonymity are improved, but system complexity increases
Solution Approach 1:
The patent segments data into multiple tables (translation table, deviation table, cluster table, reference table) where each table contains only partial information. This segmentation achieves strong anonymity and security because no single table reveals the original data, while the modular structure allows systematic management of the complexity through clear separation of concerns.
Solution Approach 2:
The patent creates a universal anonymization framework that can handle various types of data through a consistent multi-table approach. The same clustering and table generation methodology applies across different data types, providing a unified solution that manages complexity through standardization rather than requiring separate systems for different data kinds.
4Reliability
If multiple tables are stored geographically separated to increase security, then data security is improved, but data retrieval time may increase
Solution Approach 1:
The patent performs preliminary actions by pre-computing and storing data in an optimized clustered format across multiple tables before retrieval is needed. The data is organized in advance with clustering and indexing structures that enable efficient queries, so that when retrieval occurs, the system can quickly access the pre-organized data without performing complex computations in real-time, thus minimizing retrieval time despite geographic distribution.
Data Source
AI summary
A method and system for the storage of data in compliance with territorial privacy laws while retaining data sovereignty. The method generally comprises splitting input data into clusters and anonymising the clustered data. Information relating to the anonymised data clusters may then be split and stored separately in data shards. The anonymization and splitting of the data prevents the restoration of the input data without access to all of the data shards. Also a method and system for decoding data anonymised by the method.