Blurring and Bucketing

Indirect identifiers or „quasi-identifying values“ such as age and date of birth, as well as descriptors like occupation and marital status, can all used to re-identify individuals, if enough of these attributes are present in the dataset and/or they can be aggregated into a supergroup with similar values.

For this reason, your jobs in IRI FieldShield Data masking product (or the IRI Voracity ... (Data Management Platform) apply one or more additional techniques to anonymize these data values while keeping them realistic and accurate enough for research or marketing purposes. Numerical Blur functions generate random noise for specific age and date ranges. Bucketing-Features, which generalize the values into broader categories, also anonymize quasi-identifiers.

In the sample job description below, certain age groups are divided into decade groups, multiple marital status attributes are grouped into two broader categories in a defined state, educational qualifications are simplified by a new set lookup file, and all occupations have been explicitly reprocessed.

These job specifications can be automatically generated in expedient graphical assistants and function-specific dialogs. The new result set can now be through the Risk Scoring Assistant be performed again to obtain a further determination of re-identification risk based on now less pronounced quasi-identifying attributes.