Shortly after it was rst introduced in 2006, dierential privacy became the agship data privacy denition. Since then, numerous variants and extensions were proposed to adapt it to dierent scenarios and attacker models. In this work, we propose a systematic taxonomy of these variants and extensions. We list all data privacy denitions based on dierential privacy, and partition them into seven categories, depending on which aspect of the original denition is modied.
These categories act like dimensions: Variants from the same category cannot be combined, but variants from dierent categories can be combined to form new denitions. We also establish a partial ordering of relative strength between these notions by summarizing existing results. Furthermore, we list which of these denitions satisfy some desirable properties, like composition, post-processing, and convexity by either providing a novel proof or collectingexisting ones.

Offers a systematic approach to differential privacy Addresses quantification and variation of privacy loss Lists and categorises privacy definitions

Klappentext

1. Introduction.- 2. Di erential Privacy.- 3. Quanti cation of privacy loss.- 4. Neighborhood de nition (N).- 5. Variation of privacy loss (V).- 6. Background knowledge (B).- 7. Change in formalism (F).- 8. Relativization of the knowledge gain (R).- 9. Computational power (C).- 10. Summarizing table.- 11. Scope and related work.- 12. Conclusion.

   Inhalt

   1. Introduction.- 2. Dierential Privacy.- 3. Quantication of privacy loss.- 4. Neighborhood denition (N).- 5. Variation of privacy loss (V).- 6. Background knowledge (B).- 7. Change in formalism (F).- 8. Relativization of the knowledge gain (R).- 9. Computational power (C).- 10. Summarizing table.- 11. Scope and related work.- 12. Conclusion.