Database Named Entity Matching via Levenshtein and N-grams

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current methods for managing and matching named entities in databases are inefficient due to graphical and syntactic variability, relying heavily on manual processes and techniques like Levenshtein distance or phonemization, which are language-dependent and require expert intervention.

Innovation Solution

A combined method using Levenshtein distance, n-grams, and relational analysis to create a similarity measurement, merging data to construct entity classes, and applying regularized similarity to weight partitions, enabling automatic matching and optimization of database management.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual matching processes are used for named entities, then matching accuracy can be maintained through expert judgment, but productivity is severely reduced and time consumption increases

Engineering Contradiction:

Improvematching accuracyVSAvoidprocessing speed

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The system enables automatic self-matching of named entities through computational algorithms (Levenshtein distance, phonemization, n-grams) that process and match entities without human intervention, allowing the system to serve itself in the matching task while maintaining both accuracy and productivity

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

Manual expert judgment is replaced by automated computational methods including Levenshtein distance calculation, phonemization algorithms, and n-grams analysis, substituting the mechanical human matching process with automated software-based systems that operate faster and more consistently

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Ease of manufacture

If Levenshtein distance method is used alone for matching, then the process is simple to implement, but matching results are poor and reliability is reduced

Engineering Contradiction:

Improveimplementation simplicityVSAvoidmatching result quality

Core Design Contradiction:

Ease of manufactureVSReliability

Solution Approach 1:

Multiple matching methods (Levenshtein distance, phonemization, n-grams) are merged into a unified system where each method contributes different strengths, and their results are combined through voting or weighted aggregation to produce more reliable matching outcomes than any single method could achieve alone

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The matching system uses a composite approach combining multiple algorithmic 'materials' (different similarity measurement techniques) to create a stronger, more reliable matching solution, analogous to using composite materials in engineering to achieve superior properties beyond what single materials can provide

Inventive Principle:

Principle #40Composite materials

3Measurement precision

If phonemization method is used for matching, then matching accuracy improves for certain cases, but device complexity increases due to language dependency and expert intervention requirements

Engineering Contradiction:

Improvematching accuracyVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The phonemization component is designed to work across multiple languages and entity types, making the matching system universal rather than language-specific. The system can apply phonemization rules broadly to various named entities (persons, places, organizations) without requiring separate expert systems for each language or entity type

Inventive Principle:

Principle #6Universality (Multi-functionality)

4Reliability

If multiple matching methods are combined to improve reliability, then matching quality increases, but device complexity and computational requirements increase

Engineering Contradiction:

Improvematching result qualityVSAvoidsystem complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The complex matching process is segmented into distinct modular components (Levenshtein distance calculation, phonemization module, n-grams analysis, voting mechanism) that can be independently developed, tested, and maintained. Each segment handles a specific aspect of matching, reducing overall system complexity while maintaining high reliability through the coordinated operation of specialized modules

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS8117237B2Optimized method and system for managing proper names to optimize the management and interrogation of databases

Publication Date: 2012.02.14 THALES SA

AI summary

Optimized method and system making it possible to manage named entities or proper names that are presented in a different graphical or syntactic form and contained in a database (3), characterized in that it comprises, in combination, the implementation of the Levenshtein method by using a measurement of similarity from the Levenshtein method and of the n-grams method in order to find a canonic entity that can replace a set of named entities in a database.