Match and De-duplication Strategy

Match and De-duplication Strategy

Audience: Developers and Architects

Overview

This article explains how EIDR identifies duplicate content and how to correctly use the Match service.


Why De-duplication Matters

EIDR enforces:

One ID per unique object

De-duplication ensures:

  • No duplicate records

  • Consistent cross-system linking

  • Reliable identifiers


Match Service Behavior

The Match service analyzes submitted metadata and returns:

Result

Meaning

Result

Meaning

Match

Exact match found

No Match

No existing record

Candidates Found

Possible matches

Error

Invalid input


Recommended Workflow

Match -> Evaluate -> Register (if needed)

Handling Results

Match

  • Use returned ID

  • Do not register

No Match

  • Safe to register new record

Candidates Found

  • Requires manual or business rule decision

  • Do not auto-register


Common Mistakes

Skipping Match

Leads to duplicate IDs

Weak Metadata

  • Missing titles

  • Missing participants

  • Poor quality inputs

Auto-registering Candidates

Creates conflicts and cleanup work


Best Practices

  • Always match before register

  • Provide rich metadata

  • Use consistent title strategy

  • Normalize input data