Once the account holder name is retrieved from the bank, a matching score is calculated.

Overview

To determine a matching score:

  • When the account holder name is retrieved from the bank, it is compared to the name passed in the original POST /verification request.
  • A score is calculated, which may be returned in the response to GET /verification/{verificationId} endpoint. Pass the verification id returned in the POST /verification response.
  • Scores are returned in the range 0 to 100 with 0 being unmatched and 100% being treated as an exact match.

Scoring Logic

Nuapay uses a specific algorithm to determine the matching score. It is up to clients to decide on how to proceed, based on the score returned.

Scoring Examples

The following table gives some examples of how the provided name versus the account holder name returned from the bank might be handled.

Score

Result

Example

Client Decision

100

Exact Match

 

Name Provided/Retrieved

Verification Request:

John Smith

AIS Response:

John Smith

OR

Verification Request

Tommy Adams

AIS Response:

Thomas Adams

The client can be confident that there is a match where names match exactly.

In the second example, the algorithm sees “Tommy” and “Thomas” as being essentially a match so 100 is returned as the score here too.

0

No Match

 

Name Provided/Retrieved

Verification Request:

John Smith

AIS Response:

Susan Jones

The names do not match.

43

Partial Match

 

Name Provided/Retrieved

Verification Request:

John Smith

AIS Response:

John Smyth

The client will need to review; in this example, the client would treat this result as non-matching, as the surnames are different.

86

Partial Match

 

Name Provided/Retrieved

Verification Request:

Thalia Jones

AIS Response:

Natalia Jones Smith

The client will need to review; in this example, the client might decide that the similarity in names is enough to confirm a match.