The Preserve format masks data in accordance to format requirements and maintains the maximum data utility specific to a data class. 

For all the 165 pre-defined data classes in IBM Knowledge Catalog, Masking flow defines the data class's masking format to use for format-preserving tokenization and encryption.

There are five major data class format groups:

• Credit card numbers, social security numbers, and national identifiers. Masking flow preserves the national and international requirements for these data classes. For example, the first three digits of social security numbers in the United States can fall in only the ranges of (001 - 665, 667-899).

• First names, street names, and cities are masked with a default list of first names, street names, and cities that are common in the United States. For other determinate list-based data classes, such as states, countries, Masking flow has a list of all legal values.

• Email addresses. See the Advanced data settings for data classes table.

• Date, Date of birth, and Credit Card Expiration Dates are masked with date options. See the Advanced data settings for data classes table.

• Identifier data classes, such as business IDs and customer IDs are recommended to be masked with the Identifier masking.

By using the Preserve format method, some data classes might have more advanced settings, in addition to Consistency and Input Validation.