10th Estonian Winter School in Computer Science (EWSCS)
X Eesti Arvutiteaduse Talvekool
Palmse, Estonia, February 27 - March 4, 2005
under the auspices of European Educational Forum
Benny Pinkas
Trusted Systems Lab
HP Labs
Haifa, Israel
Privacy-Preserving Data Mining
Abstract
The digital age has enabled widespread access to and collection of data. While there are several advantages to ubiquitous access to data, there is also the potential for breaching the privacy of individuals.
Privacy-preserving data mining studies techniques for meeting the potentially conflicting goals of respecting individual rights and allowing legitimate organizations to collect and mine massive data sets. One approach for solving such problems is based on cryptographic protocols allowing mutually distrusting parties to evaluate a common function while leaking as little information as possible about their inputs. Another approach is based on statistical methods, for example database algorithms that compute the true response to queries and add noise to the answer, reporting only the noisy answer.
In this course we will study privacy-preserving data mining techniques, with an emphasis on using rigorous definitions and analysis. The course will begin with a brief introduction to cryptography.
Course materials
- B. Pinkas. Privacy-preserving data mining. Lecture 1: Motivation, introduction to crypto. Slides. [ppt]
- B. Pinkas. Privacy-preserving data mining. Lecture 2: Cryptographic solutions. Slides. [ppt]
- B. Pinkas. Privacy-preserving data mining. Lecture 3: Non-cryptographic approaches for preserving privacy. Slides. (Based on slides of K. Nissim.) [ppt]
About the Lecturer