Privacy Protection with Uncertainty and Indistinguishability
X. Sean Wang, Sushil Jajodia · Auerbach Publications eBooks · 2007
Contents 9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 9.2 Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 9.3 Indistinguishability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 9.4 Technical Challenges and Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 9.5 Other Related Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 9.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 9.1 Introduction In many data systems, it is important to protect individual privacy while satisfying application requirements. To provide such protection, privacy disclosure must be measured in some quantitative manner, as absolute privacy is usually not a practical proposition. Privacy measurement metrics have appeared in the literature, but they are either for single table scenarios (e.g., [17,22,23]), or for a more theoretical purpose (e.g., [20]). This chapter introduces two data privacy measures that can be used for general relational data releases and that are amenable to practical applications, and outlines challenges and possible solutions in using these measures in applications.