ML-Based Relational to Non-Relational Query Offloading

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The challenge lies in efficiently offloading data from relational databases to non-relational databases, where traditional methods struggle to retrieve specific queries due to the lack of linked tables in non-relational databases, leading to data inconsistencies and inefficient data retrieval.

Innovation Solution

A machine learning-based system that converts relational queries into non-relational queries by utilizing primary and foreign keys to create access patterns, allowing for systematic selection and validation of queries, thereby enabling efficient data retrieval in non-relational databases.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If data is offloaded from relational database to non-relational database, then storage flexibility and expansion capability are improved, but data retrieval efficiency deteriorates due to lack of linked table structure

Engineering Contradiction:

Improvestorage flexibilityVSAvoiddata retrieval efficiency

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The system performs preliminary analysis of access patterns from relational queries before converting to non-relational format. By pre-identifying frequently accessed fields and their relationships, the system creates an optimized document structure with embedded indexing information that anticipates future retrieval needs, thus maintaining efficiency after the one-way conversion.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary mapping layer that translates between relational query concepts and non-relational document structures. This intermediary maintains the logical relationships between data elements through structured field mappings and embedded reference systems, enabling efficient retrieval without requiring the original relational table structure.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Ease of manufacture

If traditional offloading methods are used without access patterns, then implementation simplicity is improved, but query accuracy deteriorates leading to incorrect data retrieval

Engineering Contradiction:

Improveimplementation simplicityVSAvoidquery accuracy

Core Design Contradiction:

Ease of manufactureVSMeasurement precision

Solution Approach 1:

The system implements self-service through automated access pattern generation that requires minimal manual configuration. The machine learning model automatically analyzes historical query patterns and generates optimized access patterns from the converted data structure, enabling the system to self-optimize without extensive manual tuning while maintaining high query accuracy.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent incorporates feedback mechanisms where query results are continuously analyzed to refine and update access patterns. The system monitors retrieval accuracy and uses this feedback to iteratively improve the mapping between relational queries and non-relational document structures, thereby maintaining high query accuracy over time.

Inventive Principle:

Principle #23Feedback

3Reliability

If machine learning process is implemented for query conversion, then data consistency is improved, but computing resources and time consumption increase

Engineering Contradiction:

Improvedata consistencyVSAvoidcomputing resources

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The system performs comprehensive data consistency validation and access pattern generation as preliminary actions during the initial conversion phase. By establishing the mapping structure and validating data integrity upfront, the system avoids the need for continuous expensive validation operations during query execution, thus reducing ongoing computing resource consumption.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent creates simplified copy structures of the relational data model in the non-relational format, maintaining only the essential relationships and data integrity constraints needed for accurate retrieval. This selective copying approach preserves data consistency while minimizing the computational overhead compared to maintaining full relational integrity checks.

Inventive Principle:

Principle #26Copying

Data Source

PatentUS10838952B2Resource offloading process using machine learning for database

Publication Date: 2020.11.17 BANK OF AMERICA CORP

AI summary

A resource offloading system comprises a relational database configured to store relational queries including searching features, a server configured to offload the relational queries to non-relational queries, and a non-relational database configured to store the non-relational queries. The server is configured to generate access patterns for each relational query based on searching features included in the relational query, classify the access pattern based on a terminology defined by a searching feature into a classification, rank the classified access patterns, generate candidate access patterns from the ranked access patterns, create non-relational queries based on the candidate access patterns which are ranked beyond a threshold, and store the non-relational queries in the non-relational database.