Blocking documentation

MATCHFORGE PLAYGROUND

Blocking comparison

Compare record matching before and after blocking. See how candidate generation reduces comparisons and how it affects the matching results.

INPUT

Two datasets

Source and target records

CANDIDATE GENERATION

Blocking

Reduce possible pairs

FINAL STEP

Similarity matching

Score and rank candidates

01

Input datasets

Both approaches use the same source and target records.

A

Source dataset

5 records · 3 columns

customer_nameemailcity

Edit cell values directly or add and remove records.

B

Target dataset

6 records · 3 columns

full_nameemail_addresslocation

Edit cell values directly or add and remove records.

02

Matching configuration

These settings remain identical in both approaches.

Field mappings

Choose which fields contribute to the similarity score.

80%
10% · More matches100% · Stricter
03

Blocking rules

These rules are applied only to the blocked approach.

OR logic: A pair is retained if it satisfies at least one rule. Duplicate candidate pairs are removed.

Normalization settings

Standardize values before generating blocking keys.

▼

Ready to compare?

Run both approaches using the same datasets and matching settings.