Relational Database Incremental Gradient Processing

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Connecting numeric analysis programs to relational database management programs is inefficient, requiring data export and re-import, which is error-prone, slow, and difficult, especially for large data sets.

Innovation Solution

Implementing incremental gradient methods directly within the relational database management program using standard database structures, allowing tuple-by-tuple application of gradient functions and dynamic termination conditions, thereby reducing unnecessary data transfer and leveraging optimization techniques.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If numeric analysis programs are connected to relational database management programs through data export and re-import, then data analysis functionality is achieved, but processing speed decreases and error probability increases

Engineering Contradiction:

Improvedata integrityVSAvoidprocessing speed

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent merges the numeric analysis functionality directly into the relational database management program by implementing gradient descent algorithms within the database engine. This allows the database to perform statistical computations natively on stored data without exporting to external programs, eliminating data transfer errors and maintaining data integrity while significantly improving processing speed.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent introduces an intermediary layer within the database management system that bridges the gap between traditional database operations and numeric analysis. This intermediary implements optimized gradient computation routines that operate directly on database tuples, serving as a mediator between the database storage structure and the statistical analysis requirements.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If data is exported piece-by-piece for execution on numeric analysis program, then complex statistical computations can be performed, but time consumption increases

Engineering Contradiction:

Improvestatistical analysis capabilityVSAvoidcomputation time

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent implements continuous gradient descent computation by processing database tuples in a continuous stream rather than exporting data in discrete pieces. The gradient computation maintains continuous state information across tuple processing, allowing the algorithm to converge efficiently without the interruptions and data reassembly required by piece-by-piece export approaches.

Inventive Principle:

Principle #20Continuity of useful action

Solution Approach 2:

The patent segments the gradient computation into tuple-level operations that can be executed independently and incrementally. Each tuple contributes to the gradient calculation in a modular fashion, allowing the computation to proceed through the database in efficient chunks without requiring complete data export, thus reducing time consumption while maintaining statistical analysis capability.

Inventive Principle:

Principle #1Segmentation

3Speed

If relational database management program enforces particular data structure for improved access speed, then data access efficiency increases, but flexibility for complex numeric operations decreases

Engineering Contradiction:

Improvedata access speedVSAvoidnumeric analysis flexibility

Core Design Contradiction:

SpeedVSAdaptability or versatility

Solution Approach 1:

The patent makes the relational database management program multi-functional by integrating numeric analysis capabilities directly into the database engine. The same data structures optimized for fast access are simultaneously used for both traditional database operations and gradient-based statistical computations, eliminating the need for separate optimized structures while maintaining both speed and flexibility.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS9208203B2High-speed statistical processing in a database

Publication Date: 2015.12.08 RE CHRISTOPHER M

AI summary

Numerically intensive statistical processing of data is implemented as an incremental gradient method within the engine of a database system. Small user-defined functions in the database system calculate an approximate gradient from one term of a linearly separable defined cost resolvable from a single tuple of the database. In this way the optimized data access of the database may be exploited for rapid statistical processing.