Security, Privacy, and Economic Incentives

1
Security, Privacy, and Economic Incentives

• Sheng Zhong

2
Personal Background

• Ph.D., Yale University (2004)
• 1-Year Postdoc at DIMACS (Discrete
• Math and Theoretical Computer Science)
• Joined UB CSE in Fall 05
• Received NSF research grant (200K) in
• September 05
• Doing research on computer security, data
  privacy, and economic incentives

3
Why Security, Privacy Incentives?

• The need to fight non-cooperative behavior in
  computer systems.
• To fight really bad guys Security
• To protect every individual Privacy
• To promote good behavior among rational people
  Economic Incentives

4
Lines of Research

• Database Privacy
• Privacy in Wireless Networks
• Applied Cryptography
• Todays Presentation
• 1. Privacy-Preserving Data Mining
• 2. Economic Incentives in Mobile Computing (and
  more)

5
1 Privacy-Preserving Data Mining (PPDM)

• Data Mining Discovering rules/patterns in huge
  amounts of data.
• Example Finding relationship between illness
  lifestyle from Internet survey.
• Privacy Concerns Prevent privacy violation in
  data mining.
• Example Do I have to reveal my medical history
  in the above survey?

6
A Scenario of Data Mining with Privacy Concern
?
Private Information
Personal Data
7
Current Solutions to PPDM
8
Our Work on PPDM

• All on customer-miner scenario.
• Faster crypto solution
• Privacy-preserving data collection ? generic
  crypto solution.
• Accurate perturbation solution.

9
2 Economic Incentives

• Consider a protocol (in mobile computing,
  distributed data mining, etc.)
• Good/honest participants follow it and do nothing
  else
• Bad/malicious participants deviate from it and do
  arbitrary things
• But very few people are completely honest or
  malicious most of them are economically
  rational.
• They do whatever is the best for themselves.

10
Example Incentives in Mobile Computing

• Ad hoc network multi-hop mobile network without
  pre-existing infrastructure.
• Nodes depend on other nodes to relay packets but
  other nodes have no incentive to do so.

packet
11
Our Work on Incentive-Compatible Protocol

• Our Work 1 Simple game theoretic solution to
  packet forwarding game.
• Our Work 2 General framework for
  routingforwarding.
• Using game theory cryptography.

12
New Project Incentive-Compatible Protocols

• Newly funded by NSF Cyber Trust program, Sep
  2005-Aug 2008
• Studies incentives not only in mobile computing,
  but also in distributed data mining.
• Also investigates fundamental theory and
  experimental tools of general incentive-compatible
  protocols.
• Will be a focus of my future research.

13
Summary Selected Publications (1)

• On Designing Incentive-Compatible Routing and
  Forwarding Protocols in Wireless Ad-Hoc Networks.
  Sheng Zhong, Li Erran Li, Yanbin Grace Liu, and
  Yang Richard Yang. In MOBICOM 2005 (acceptance
  ratio 10.3)
• Anonymity-Preserving Data Collection. Zhiqiang
  Yang, Sheng Zhong, and Rebecca N. Wright. In KDD
  2005 (full paper, acceptance ratio 12).
• Privacy-Enhancing k-Anonymization of Customer
  Data. Sheng Zhong, Zhiqiang Yang, and Rebecca N.
  Wright. In PODS 2005 (acceptance ratio 20.4).
• Verifiable Distributed Oblivious Transfer and
  Mobile Agent Security. Sheng Zhong and Yang
  Richard Yang. To appear in ACM Mobile Networks
  and Applications special issue (acceptance ratio
  19).

14
Selected Publications (2)

• Privacy-Preserving Classification without Loss of
  Accuracy. Zhiqiang Yang, Sheng Zhong, and Rebecca
  N. Wright. In SDM 2005 (full paper, acceptance
  ratio 18.3).
• Towards a Theory of Data Entanglement. James
  Aspnes, Joan Feigenbaum, Aleksandr Yampolskiy,
  and Sheng Zhong. In ESORICS 2004 (acceptance
  ratio 17).
• Sprite A Simple, Cheat-Proof, Credit-Based
  System for Mobile Ad-Hoc Networks. Sheng Zhong,
  Jiang Chen, and Yang Richard Yang. In INFOCOM
  2003 (acceptance ratio 20.8).
• Optimistic Mixing for Exit-Polls. Philippe Golle,
  Sheng Zhong, Dan Boneh, Markus Jakobsson, and Ari
  Juels. In ASIACRYPT 2002 (acceptance ratio
  19.7).

15
End of Talk

• Thank You!
• Questions?