Advanced anonymization

In this section, we will explore how you can apply anonymization to more complex scenarios and with measurable anonymization, which aims to address linkage attacks by introducing degrees of anonymity reflecting linkage. We will start by looking at k-anonymity to achieve measurable anonymity using techniques such as clustering.

K-anonymity

This technique anonymizes data based on the degree of anonymity (k), which ensures that each record in a dataset is indistinguishable from at least k−1 other records concerning certain identifying attributes. The approach aims to make linking data to an individual harder. For instance, in our previous simple masking example, we masked the last three characters of the postcode. With k-linking, the number of masked attributes would depend on how many individuals could be identified with the postcode and all the other attributes.

You can find an implementation in the k-anonymity library at https://github.com/PacktPublishing...