PingOne

PingOne Verify and identity data matching

Identity data matching in PingOne Verify manages how identity data is captured and presented across different document types to allow organizations to customize their verification criteria and meet their risk and user experience goals.

How identity data matching works

Identity data matching compares user-provided identity attributes, such as name and date of birth, against attributes extracted from an identity document. The structure and encoding of these attributes vary by document type.

For example, U.S. driver licenses typically encode and display names using two fields:

  • Given Name: Includes both the first and middle names.

  • Last Name: The surname.

During the matching process, PingOne Verify treats these components independently.

Scoring and configuration

If only a portion of a full name in a U.S. driver license is provided during verification (first name without the middle name), PingOne Verify can assign a medium confidence score for that attribute because of the partial match.

A common identity data match configuration that’s widely used is:

  • Given Name: Medium confidence allowed

  • Last Name: High confidence required

  • Date of Birth: High confidence required

This configuration provides strong identity assurance and flexibility for users who might not provide their full given name.

Handling middle names and nicknames

PingOne Verify identity data matching also supports scenarios where users go by a middle name or nickname, which can differ from the first name on their ID.

If the provided name differs from the official first name but matches the middle name, PingOne Verify compares the full given name on the document against the input. This often results in a medium score rather than a full mismatch.

The AI-powered engine also recognizes common nickname variations across different languages and cultures. For example:

  • Kenny: A shortened form for Kenneth

  • Kay: A shortened form of Katherine

The matching algorithm incorporates these mappings into the score to improve match tolerance and user pass rates, while still protecting against fraud.

Handling identity documents without name delimiters

Sometimes certain identity documents, especially international ones, don’t include clear line breaks or structured delimiters between name components.

When names are presented as a single unbroken string, PingOne Verify extracts the full name without a reliable way to distinguish between given names and last names.

In these cases:

  • Only the fullName attribute is returned from the document.

  • Attempts to match against separate givenName or lastName attributes can result in a mismatch because there’s no structural separation in the source data.

To ensure accurate results, you should:

  • Provide the full expected name (given name, last name) for matching against these document types.

  • Configure identity data matching rules to use the Full Name field for comparison. Learn more in Creating a verify policy.

  • Adjust scoring configurations in the PingOne admin console Verify Policies page to evaluate the fullName attribute directly rather than relying on split name fields.

This approach ensures compatibility with a broader set of document formats and improves match rates when field separation is unavailable.

AI-enhanced identity data matching

PingOne Verify uses AI models trained on global naming conventions and linguistic patterns to enhance the accuracy and flexibility of identity data matching.

This enables identity data matching to:

  • Account for partial names

  • Detect common nickname-to-full-name relationships

  • Maintain a high-level confidence even in the presence of name variations