Probabilistic Index Update for Database Retrieval Efficiency

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The existing methods for managing databases with indices face inefficiencies, particularly when data is updated frequently, leading to excessive processing loads and variable response times during retrieval requests due to frequent index updates.

Innovation Solution

A method where index updates are performed with a determined probability at data update time and also before data retrieval, using a combination of eager and lazy update approaches to control the frequency of index updates and maintain efficient data retrieval performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the index is updated every time data is updated, then the retrieval performance is improved, but the processing load on the system becomes excessive

Engineering Contradiction:

Improveretrieval performanceVSAvoidprocessing load

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

Instead of updating the index after every data update (excessive action), the patent applies partial action by updating the index only when a random number generated during data update is less than a predetermined threshold probability. This selective updating approach reduces the processing load while maintaining adequate retrieval performance.

Inventive Principle:

Principle #16Partial or excessive action

2Productivity

If the index is not updated at record update time, then the processing efficiency is improved, but the response time for retrieval requests varies and becomes longer after large data updates

Engineering Contradiction:

Improveprocessing efficiencyVSAvoidresponse time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The patent implements preliminary action by generating a random number during the data update process itself (rather than waiting for retrieval time) to determine whether to perform index update. This preliminary decision-making reduces the response time variability by avoiding the need to evaluate update necessity at retrieval time.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses feedback from the random number generation during data update to dynamically control index update behavior. The predetermined threshold probability acts as a feedback mechanism that balances processing efficiency with response time requirements.

Inventive Principle:

Principle #23Feedback

3Measurement precision

If the index structure is rebuilt frequently to maintain accuracy, then the data retrieval accuracy is improved, but the system processing load increases excessively

Engineering Contradiction:

Improvedata retrieval accuracyVSAvoidsystem processing load

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent changes the parameter of index update frequency from deterministic (every update) to probabilistic (based on random number and threshold). This parameter change allows the system to maintain adequate retrieval accuracy while significantly reducing the processing load on the system.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS10437806B2Database management method and information processing apparatus

Publication Date: 2019.10.08 FUJITSU LTD

AI summary

An information processing apparatus includes a determination unit and an output unit. If data included in a database with an index is updated, the determination unit determines whether to perform an index update process so that the index update process will be performed with a determined probability. If the determination unit determines to perform the index update process, the output unit outputs an index update request.