'What is meant by k-anonymity and l-diversity, and what is difference between them? [closed]
What is meant by k-anonymity and l-diversity, and what is difference between them? This concept is used in the following IEEE paper.
Solution 1:[1]
an anonymised dataset contains multiple fields that can be used to identify someone (eg. age, sex, location). even if one field is not unique, a combination of them may be (eg. there may be only 1 male aged 54 in squatsville). k-anonymity greater than 1 means that even with all fields you don't identify a single person - you identify "k" people.
but that's not enough, because all k people identified by a distinct set of the different fields might be associated with the same value - if there are 10 different 54 year-old males in squatsville and they all wear ladies clothes, then you know that bob, who is 54 and lives in squatsville, wears ladies clothes. so you need diversity within the group identified by the fields. l-diversity means that the group contains "l" (the letter L, not the digit 1) different values.
[no offense intended - seriously - to cross-dressers. i just wanted something mildly amusing that wasn't, well, cancer...]
and a reference - http://www.cs.colostate.edu/~cs656/reading/ldiversity.pdf
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|---|
| Solution 1 |