Identity Verification via Fuzzy Match Scoring

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing identity verification systems fail to accurately verify entities when there are inconsistencies in data, leading to multiple attempts and time-consuming searches across data sources.

Innovation Solution

An identity verification system generates a search structure with alternative queries based on subject attribute data, using wildcard tokens and character substitutions to accommodate variations, and determines a match score for candidate records to select the best match, even in the presence of errors or format differences.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If exact match search is performed against data sources, then verification accuracy is improved, but verification fails when data inconsistencies exist leading to multiple attempts

Engineering Contradiction:
Improveverification accuracyVSAvoidverification success rate
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent changes the search parameters from exact match to fuzzy match by introducing match score thresholds and similarity calculations. The system calculates match scores between candidate records and subject attribute data, allowing verification to succeed even when data inconsistencies exist, thereby resolving the contradiction between verification accuracy and verification success rate

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent applies partial matching instead of requiring complete exact matches. By allowing partial matches above a certain threshold and ranking candidates by match score, the system can verify identities even with incomplete or slightly inconsistent data, improving verification success rate while maintaining reasonable accuracy

Inventive Principle:
Principle #16Partial or excessive action

2Reliability

If multiple verification attempts are performed to handle inconsistent data, then verification completeness is improved, but time consumption increases

Engineering Contradiction:
Improveverification completenessVSAvoidverification time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent performs preliminary fuzzy matching and ranking before final verification decisions. By pre-calculating match scores and ranking candidate records, the system identifies the most likely matches early, reducing the need for multiple verification attempts and thereby reducing time consumption while maintaining verification completeness

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent replaces the mechanical process of multiple sequential verification attempts with an automated fuzzy matching and ranking system. The match score calculation and candidate ranking automatically handle inconsistent data without requiring manual retry cycles, significantly reducing verification time while maintaining completeness

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Adaptability or versatility

If fuzzy matching with wildcards is used to handle data variations, then search coverage is improved, but false positive matches increase

Engineering Contradiction:
Improvesearch coverageVSAvoidmatch accuracy
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent uses match score feedback to filter and rank fuzzy match results. By calculating and comparing match scores against thresholds, the system can adjust the balance between search coverage and match accuracy, allowing broad search coverage while filtering out false positives through score-based ranking and selection

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP2866156B1Method and apparatus for identity verification
Publication Date: 2015.09.02 GB GRP PLC
  • EP2866156B1 patent drawingFigure 1~5
  • EP2866156B1 patent drawingFigure 2
  • EP2866156B1 patent drawingFigure 3

AI summary

Embodiments of the present invention provide a computer-implemented method of verifying an identity of an entity against one or more data sources (131, 132, 133) wherein a record value is determined for each of a plurality of candidate records, determined based a score of each item of candidate attribute data associated with the record.