Adaptive Data Protection via ML Classification and Blockchain Tokenization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data storage systems are vulnerable to data leakage due to endpoint security breaches, as they rely on manual classification and encryption methods that are inefficient and prone to human error, and do not effectively protect multi-tenant data across unencrypted network channels.
Innovation Solution
An adaptive distributed data protection system utilizing machine learning and blockchain technology to identify and categorize sensitive data, apply token-based encryption, and store it on a secure ledger, ensuring only authorized tenants can access their data, thereby enhancing security and compliance with regulatory standards.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is distributed across multiple compute/storage nodes, then security against endpoint breaches is improved, but system complexity increases
Solution Approach 1:
The patent divides data into multiple fragments and distributes them across different compute/storage nodes. Each node stores only a portion of the encrypted data, so that compromising a single endpoint does not expose the complete dataset. This segmentation approach directly resolves the contradiction by enhancing security through distribution while managing complexity through systematic data fragmentation and encryption protocols.
2Reliability
If manual classification and identification of sensitive data is performed, then data protection coverage is improved, but efficiency deteriorates and human error increases
Solution Approach 1:
The system implements automated classification and identification mechanisms that operate without human intervention. The patent employs cryptographic protocols and system-generated metadata to automatically identify and classify sensitive data, eliminating the need for manual data labeling. This self-service approach maintains comprehensive data protection coverage while dramatically improving efficiency and eliminating human error associated with manual classification.
3Ease of operation
If static user-defined protection definitions are used, then ease of operation is improved, but adaptability to new security threats deteriorates
Solution Approach 1:
The patent implements dynamic protection definitions that automatically adapt to new security threats and data types. The system uses cryptographic protocols that can be updated and reconfigured without requiring changes to the underlying infrastructure or user reconfiguration. This dynamic approach maintains ease of operation while significantly improving adaptability, as the system can respond to emerging threats through automated protocol updates rather than requiring manual redefinition of protection rules.
Data Source
AI summary
Embodiments for protecting data stored and transmitted in a computer network, by receiving confidential data from a client, the data organized into labeled fields and corresponding data elements; filtering the received data to identify fields that require data masking; generating a security prediction on the corresponding data elements using a machine learning process; separating the masked data into tokenized data having a respective token associated with each corresponding data element; and storing the tokenized data on a blockchain secure ledger to ensure integrity of the received data and prevent an ability to tamper with the received data.