Data pseudonymization
De-identify and mix names or nouns with realism
Challenges
When masking data or generating useful test data, you need output values that look real but do not contain personally identifiable information (PII). This is especially true for the names of people, places, and things.
Encryption, de-identification, and other masking, hashing, and bit obfuscation functions protect sensitive data but don’t provide the level of realism certain recipients need. They need a simpler way to alter the individualization characteristics of data with a substituted, but realistic, output value.
You must also ensure that the real name cannot be easily discovered through reversal or guesswork.
Solutions
When working with PII in tables or flat files, use IRI FieldShield – or the SortCL-program in the IRI CoSort Product or in the IRI Voracity-Platform – to replace this data with secure but realistic dummy outputs, which are stored in DB tables or external datasets, so-called Set files, are saved. If you need to do the same with ranges in Excel, use IRI CellShield, or for unstructured data sources IRI DarkShield. They support:
|
Recoverable pseudonymization |
Specify a lookup set where real and fake names are either pre-filled or randomly assigned. Use the recovery set to restore the original names. |
|
Irreversible pseudonymization |
Select random replacement names for the original value from a set file containing real or fake names. This way, the original name value has no automatic basis for recovery. |
Set the pseudonym method used in your output fields in simple 4GL job scripts, or use the Pseudonymization dialog in the masking dialogs of FieldShield-GUI or the DarkShield Assistant in the same Eclipse™ IDE or in CellShield, which also supports pseudonymized lookup replacements of values in Excel.
Pseudonymization is just one method by which FieldShield can identify information in a dataset. It can also pseudonymize other Functions combine data security at the field level.
Do you need test names?
In addition to pseudonymization and other masking of production data, there is a standalone solution for creating secure but realistic first and last names of genders (or other nouns). IRI RowGen uses the same metadata as FieldShield (and SortCL) to create and format pseudonyms for use as test data values (or in formatted test data destinations).
RowGen is particularly necessary to provide anonymous yet realistic test data when production data is unavailable or insufficient. RowGen builds structurally and referentially correct test data into database, file, and report targets. Note that RowGen is also included in Voracity.
Related solutions
Frequently Asked Questions