Hierarchical Data Spaces for Encrypted Operations Without Full Scans
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Relational and non-relational data management systems face inefficiencies in data operations due to the need to scan entire columns of unsorted values, exposing data and increasing inference risks when values are not encrypted.
Innovation Solution
Implementing a system that uses logical hierarchical data spaces (LHDS) to organize and store data, allowing probabilistic operations through hierarchical path identifiers (HPId) that preserve data hierarchy, enabling encryption and efficient access without comparing values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If relational data management systems use column-oriented approaches to store and retrieve information by comparing or sorting values, then data operations can be performed by value for relevant attributes, but the system must scan the entire column of values when columns are unsorted, increasing time complexity and exposing data to inference risks
Solution Approach 1:
The patent segments data into hierarchical levels (root level and child levels) organized by attribute groups. Each level contains subsets of data that are sorted and indexed separately, allowing the system to scan only relevant segments rather than entire columns. This segmentation maintains security by encrypting data at each level while enabling efficient access through hierarchical indexing.
Solution Approach 2:
The patent introduces a hierarchical dimension to traditional flat column storage. Instead of storing all values in a single unsorted column, the system organizes data across multiple hierarchical levels with different sorting criteria. This dimensional transformation allows the system to navigate through encrypted data structures using hierarchical path identifiers (HPId) without decrypting entire columns, thus maintaining security while reducing scan time.
2Productivity
If data is stored in unsorted columns to simplify storage structure, then device complexity is reduced, but the system must scan entire columns to perform operations, reducing productivity
Solution Approach 1:
The patent divides data into segments organized by attribute groups at different hierarchical levels. Each segment is sorted and indexed independently, allowing efficient data operations without requiring the entire column to be sorted. This segmentation approach improves productivity by enabling selective scanning of relevant segments while maintaining manageable complexity through structured organization.
Solution Approach 2:
The patent performs preliminary sorting and indexing of data at each hierarchical level during the data loading phase. This preliminary action creates sorted segments and indexes that can be quickly accessed during operations, eliminating the need for full column scans during query execution. The preliminary organization of data structures enables efficient retrieval without requiring complex runtime sorting operations.
3Reliability
If data is encrypted to protect against inference risks, then data security is improved, but the system cannot efficiently compare or sort values without decrypting, increasing computational overhead
Solution Approach 1:
The patent introduces hierarchical path identifiers (HPId) as intermediaries that represent encrypted data positions and relationships. Instead of decrypting data to compare values, the system operates on HPId structures that preserve hierarchical relationships. This intermediary mechanism allows efficient comparison and sorting operations on encrypted data by manipulating the HPId representations rather than the actual encrypted values, thus maintaining security while reducing computational overhead.
Solution Approach 2:
The patent replaces traditional value-based comparison mechanisms with hierarchical structure-based operations. Instead of decrypting and comparing numerical or string values directly, the system uses hierarchical path identifiers to navigate and compare encrypted data structures. This substitution of mechanical comparison operations with hierarchical navigation operations enables efficient data operations on encrypted data without the computational burden of decryption and re-encryption cycles.
Data Source
AI summary
A method for storing or transmitting information may include determining at least one subdivision of at least one logical hierarchical data space. The at least one logical hierarchical data space may have a plurality of subdivisions. The method may further include creating at least one data block. The method may further include writing the at least one tuple to the at least one data block. The at least one tuple may be associated with the subdivision of the at least one logical hierarchical data space. The method may further include associating the at least one data block with the at least one subdivision of the at least one logical hierarchical data space.