Foundations of Privacy - Spring 2012

Instructor: Moni Naor

When:     Sunday 16:00--18:00
Where:    Ziskind 1
First meeting:   March 11th 2012

         Lectures           Homework         


DESCRIPTION:  The availability of fast and cheap computers coupled with massive storage devices has enabled the collection and mining of data on a scale previously unimaginable. This opens the door to potential abuse regarding individuals' information. There has been considerable research exploring the tension between utility and privacy in this context.

The goal of this course is to explore techniques and issues related to data privacy. In particular to study:

PREREQUISITES: Students are expected to be familiar with algorithms, data structures, probability theory, and linear algebra, at an undergraduate level. No prior cryptography course will be assumed.

REQUIREMENTS: Students are required to present one set of the papers and the background leading to it, write a summary as well as attend all meeting, read all assigned papers, and participate in class discussion. In addition there will be a few homework sets. You may discuss the problems with other students, but the write-up should be individual.


Lectures (last time):

Lecture 1 (March 15 2010)   Slides

Administrivia, Cryptography and Privacy, Attacks on Privacy, Impossibility of achieving Privacy in with general auxiliary information.

see Lindell's writeup for a brief intro on SFE

Narayanan Shmatikov on breaking the Netflix challenge

Lecture 2 (March 22 2010)  Slides

Finish Impossibility of achieving Privacy in with general auxiliary information, differential privacy

Dwork-Naor paper on impossibility

Shaltiel's survey on extractors

Dwork, McSherry, Nissim and Smith  Differential Privacy

Lecture 3 (April 12 2010)  Slides

Randomized Response, examples of noisy response: histograms, k-means. Global and Local Sensitivity.

Nissim, Raskhodnikova and Smith, Smooth Sensitivity and Sampling in Private Data Analysis

Blum, Dwork, McSherry and Nissim, Practical Privacy: The SuLQ Framework

Lecture 4 (April 26 2010)  Slides

The exponential mechanism, counting queries and the Blum Ligett Roth algorithm

McSherry and Talwar, Mechanism Design via Differential Privacy

Blum, Ligett and Roth, A Learning Theory Approach to Non-Interactive Database Privacy.

Lecture 5 (May 3 2010)  Slides

Answering predicate queries efficiently

Cryptographic lower bounds

Dwork, Naor, Reingold, Rothblum, and  Vadhan, On the Complexity of Differentially Private Data Release

Lecture 6 (May 10 2010)  Slides

Cryptographic lower bounds and Tracing Traitors

Interactive mechanisms

Dwork, Naor, Reingold, Rothblum, and  Vadhan, On the Complexity of Differentially Private Data Release

Roth and Roughgarden, Interactive Privacy via the Median Mechanism

Lecture 7 (May 17 2010)  Slides

Interactive mechanisms

Changing Data: Continual Output

Roth and Roughgarden, Interactive Privacy via the Median Mechanism

Lecture 8 (May 24 2010)  Slides

Changing Data: Continual Output and Pan Privacy

Dwork, Naor, Pitassi and Rothblum, Differential Privacy Under Continual Observation

Dwork, Naor, Pitassi, Rothblum, and  Yekhanin, Pan-Private Streaming Algorithms

Lecture 9 (May 31 2010)  Slides

History-Independent Hashing Schemes (and applications)
 

Naor and  Teague Anti-persistence: History-Independent Data Structures
Naor, Segev and Wieder,  History-Independent Cuckoo Hashing
Moran, Naor and Segev, Deterministic History-Independent Strategies for Storing Information on Write-Once Memories

Lecture 10 (June 7 2010)  Slides

Changing Data: Continual Output and Pan Privacy
 

Dwork, Naor, Pitassi and Rothblum, Differential Privacy Under Continual Observation

Dwork, Naor, Pitassi, Rothblum, and  Yekhanin, Pan-Private Streaming Algorithms

Lecture 11 (June 14 2010)  Slides

 

 

 

Reconstructing databases from counting queries
 

Dinur and Nissim, Revealing information while preserving privacy

Dwork and Yekhanin, New Efficient Attacks on Statistical Disclosure Control Mechanisms

Student presentations:

o    Guy Katz: RFID

o    Ilan Ko: Deanonyming network

 

Narayanan and Shmatikov. De-anonymizing Social Networks

Lecture 12 (June 21 2010)  Slides

 

 

 

Reconstructing databases from counting queries.

Private Learning

Private Approximation
 

Dinur and Nissim, Revealing information while preserving privacy

Dwork and Yekhanin, New Efficient Attacks on Statistical Disclosure Control Mechanisms

Kasiviswanath, Lee, Nissim, Raskhodnikova and Smith,  What Can We Learn Privately? 

Gupta, Ligett, McSherry, Roth and Talwar, Differentially Private Approximation Algorithms
 

Lecture 13 (June 28 2010)  Slides

 

 

 

Linear Programming Decoding

(e,d)-Differential Privacy

Computational Differential Privacy


 

Dwork, McSherry  and Talwar,  The price of privacy and the limits of LP decoding

Hardt and Talwar, On the Geometry of Differential Privacy


 

Student presentations:

o    Tomer Ashur: Collection and aggregation of IP addresses over the Internet

o    Itay Gonshorovitz: Privacy and Targeted Online Advertising

 

 

Toubiana, Narayanan, BonehNissenbaum and BarocasAdnostic: Privacy Preserving Targeted Advertising

Background Material

Systems:

Fiction

Blogs related to privacy

Advice on giving Academic Talks

Below is a compilation of source on giving talks. Some of it humorous and some contradicts each other