Matching configurations

Learn how matching rules ensure accuracy.

This section indicates the decision-making rules (matching configurations) for Global Identity features, under both the 1x1 and 2x2 verification methods. These rules have been pre-set to deliver the highest user conversion rates while maintaining strong levels of data verification security.

The rules are standard for all Sumsub clients and not available for customization.


Global Identity Verification

The Global Identity Verification (1x1) Configuration requires matching the applicable dataset within a single data source to receive a Full Match on the complete user identity profile*

Verification Result

Name Category

Address Categoryy

Date of Birth Category

ID Category

Full Match
Full Match
Any Value
Any Value
Full Match
Partial Match
Partial Match
Any Value
Any Value
Full Match
No Match

All other combinations

* The applicable country sources are screened until a Full Match on the complete identity profile is discovered. If a Full Match is not found, the system then repeatedly screens those sources for a Partial Match on the identity profile.

The Global Identity Verification (2x2) Configuration requires matching two (identical or different) applicable datasets within two data sources to receive a Full Match on the complete user identity profile*

Verification Result

1st Data Source

2nd Data Source

Full Match
Name
Full Match
+ National ID
Full Match
Name
Full Match
+ National ID
Full Match
Name
Full Match
+ National ID
Full Match
Name
Partial Match
+ National ID
Full Match
Name
Full Match
+ National ID
Full Match
Name
Full Match
+ Date of Birth
Full Match
Partial Match
Name
Full Match
+ National ID
Full Match
Name
Partial Match
+ National ID
Full Match
Name
Full Match
+ Date of Birth
Full Match
No Match

All other combinations

* The applicable country sources are screened until a Full Match on the complete identity profile is discovered. If a Full Match is not found, the system then repeatedly screens those sources for a Partial Match on the identity profile.

Data category and attribute matching

The following are case scenarios which result in a Full Match regarding Name, ID, and Address data categories *.

Name category matching scenarios

A Name Full Match is considered when any of the following scenarios is met:

  • First Name + Last Name
  • First Initial + Last Name
  • First Name + Maternal Name

Address category matching scenarios

An Address Full Match is considered when any of the following scenarios is met. Please note that certain Address attributes can be specific to regions/countries:

  • Building Number + City
  • Street + City
  • State + City
  • Building Number + Postal Code
  • Street + Postal Code
  • State + Postal Code

ID category matching scenarios

An ID category Full Match is considered when any of the following scenarios is met:

  • National ID
  • Date of Birth
* A Partial data category Match is considered such when a Match is discovered on either one of the two data attributes from the case scenarios above. A Partial ID category Match is not possible. A Partial Match in the Name category based solely on the First Initial data attribute is not accepted.

Fuzzy matching for individual attributes

These are the criteria for fuzzy matching against various attributes in the Sumsub 2x2 Address Verification Configuration when we receive records back from the data sources:

  • Name Attributes

    For a Name category attribute to generate a Match, it should be within 70% of the Levenshtein character similarity in comparison to the record returned by the data sources. E.g., if the Name attribute input is ‘Christophel’ and the actual name is ‘Christopher’, we generate a Match based on our fuzzy matching logic. This system is applicable to the first_name and last_name Name category attributes.

  • Address Attributes

    For Address category attributes like street, city, state, and postal_code, the acceptable Levenshtein character similarity percentage for a Match is also 70% in comparison to the record returned by the data sources. E.g., an Address attribute input of ‘200 Kingslee Court’ would Match against a record given by the data source for ‘200 Kingsley Court’.

  • ID Attributes

    Sumsub does not employ any kind of fuzzy matching on any data attributes within the ID data category (National ID, Date of Birth).

Examples

These are a few examples of inputs that would satisfy our criteria of fuzzy matching logic and generate a match.

Data Attribute

Input

Data Source Record

Result

first_name

Jeanette

Jeanette

Match

first_name

Jeanotte

Jeanette

Match

first_name

Gene

Jeanette

No Match

last_name

Richardson

Richardson

Match

last_name

Richardsen

Richardson

Match

last_name

Richarliset

Richardson

No Match

street

Brigadoon Drive

Brigadoon Drive

Match

street

Brigadier Street

Brigadoon Drive

No Match

city

Redwood City

Brigadoon Drive

Match

city

Renwood City

Brigadoon Drive

Match

city

Redweed County

Brigadoon Drive

No Match

state

Paris

Paris

Match

state

Parip

Paris

Match

state

Perip

Paris

No Match