Fuzzy Matching for In-Memory Database Search Recall

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional in-memory databases struggle with non-exclusionary searching, often excluding records with missing or undefined fields due to differing schemata, leading to reduced recall capabilities in search results.

Innovation Solution

Implementing a system with fuzzy indexing, fuzzy matching, and scoring algorithms that allow for separate query and scoring algorithms to operate on different fields, enabling the inclusion of records with missing fields in search results and assigning penalties for missing data, while using a distributed architecture with modules like search managers, conductors, and analytics agents to manage and process queries across clusters.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional exact matching search is used, then search speed and precision are improved, but recall capability deteriorates due to exclusion of records with missing or undefined fields

Engineering Contradiction:

Improvesearch precisionVSAvoidrecall capability

Core Design Contradiction:

Measurement precisionVSLoss of information

Solution Approach 1:

The patent changes the matching parameter from exact match to fuzzy match with configurable similarity thresholds. The system calculates similarity scores between query terms and record fields using fuzzy matching algorithms, allowing records with missing or partially matching fields to be included based on their similarity score rather than being excluded by strict equality checks.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent introduces a fuzzy matching layer as an intermediary between the query and the database records. This intermediary computes similarity scores and ranks results, acting as a buffer that translates exact query terms into flexible matching criteria, thereby recovering information that would otherwise be lost through exact matching exclusions.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If separate query and scoring algorithms are implemented, then adaptability and versatility are improved, but device complexity increases

Engineering Contradiction:

Improvesearch algorithm flexibilityVSAvoidsystem architecture complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent segments the search process into distinct modular components: query parsing, fuzzy matching, scoring, and ranking. Each component is implemented as a separate algorithmic module that can be independently configured and optimized. This segmentation allows different query algorithms and scoring functions to be combined flexibly without requiring complete system redesign.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent creates a universal search framework that can accommodate multiple query algorithms and scoring functions through a common architecture. The system uses standardized interfaces and configurable parameters that allow different algorithms to operate within the same framework, reducing the need for separate specialized systems for each algorithm type.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS9916368B2Non-exclusionary search within in-memory databases

Publication Date: 2018.03.13 FINCH COMPUTING LLC

AI summary

Methods for non-exclusionary searching within clustered in-memory databases are disclosed. The non-exclusionary search methods may allow the execution of searches where the results may include records where fields specified in the query are not populated or defined. The disclosed methods include the application of fuzzy matching and scoring algorithms, which enables the system to search, score and compare records with different schemata. This may significantly improve the recall of relevant records.