olemus
privaatsuse kontekstis: l-hajutuse nõrkuse ärakasutus, juhul kui tundliku atribuudi tõenäosusjaotus ekvivalentsiklassis erineb kogu andmekogumi omast ja võimaldab statistilist tuletust
= in the context of privacy: exploiting the vulnerability of the l-dispersity, in case where the probability distribution of a sensitive attribute in the equivalence class differs from that of the entire data set and allows statistical derivation

näiteid
https://slideplayer.com/slide/7820325/25/images/20/Skewness+attack+Prob+of+cancer+in+the+original+table+is+low.jpg

https://slideplayer.com/slide/6027080/20/images/30/Skewness+Attack+Example:+sensitive+attribute+is+HIV++(1%25)+or+HIV-+(99%25).jpg

ülevaateid
https://vixra.org/pdf/1207.0101v1.pdf

https://www.ics.uci.edu/~projects/295d/presentations/295d-tcloseness (page 6)