Campus Event Calendar
Event Entry
What and Who
Improving the Privacy of Online Social Networks and Cloud Computing
Stefan Saroiu
Microsoft Research
SWS Colloquium
Stefan Saroiu is a researcher in the Networking Research Group at Microsoft Research in Redmond. Stefan's research
interests span large-scale distributed systems, mobile systems, and computer security. Before coming to MSR in 2008,
Stefan spent three years writing papers, teaching, and advising students, which is pretty much what the job of an
Assistant Professor at the University of Toronto is. Before that, Stefan spent four months at Amazon.com measuring their
workload and participating in the early stages of the design of their new shopping cart service (aka Dynamo). Stefan
finished his Ph.D. in 2004 at the University of Washington where he was advised by Steve Gribble and Hank Levy.
SWS, RG1
Expert Audience
English
Note: We use this to send email in the morning.
Date, Time and Location
Friday, 4 December 2009
14:00
60 Minutes
E1 5
5th floor
Saarbrücken
Abstract
This talk has two parts. In the first part, I will present Lockr, a system that improves the privacy of online social
networks (OSNs). Lockr offers three significant privacy benefits to OSN users. First, it separates social networking
content from all other functionality that OSNs provide. This decoupling puts users in control of their own social
information: they decide which OSN providers should store it, which third parties should have access to it, or they can
even choose to manage it themselves. Second, Lockr ensures that digitally signed social relationships needed to access
social data cannot be re-used by an OSN site for unintended purposes. This feature drastically reduces the value of the
social content that users entrust to their OSN providers. Finally, Lockr enables message encryption using a social
relationship key. This key lets two strangers with a common friend verify their relationship without exposing it to
others, a common privacy threat when sharing data in a decentralized scenario.
In the second part, I will present our ongoing work in improving privacy when users run their code in the
infrastructure. This is becoming an increasingly common scenario as cloud computing and mobile computing are becoming
more popular. I will start by discussing these scenarios, after which I will present our current work in minimizing the
attack surface exposed to malicious users and operators in VM-based cloud environments.
Lockr is joint work with Alec Wolman (MSR), Amin Tootoonchian, and Yashar Ganjali (U. of Toronto). For more information
or to download our Lockr implementations for Flickr and for BitTorrent, please visit http://www.lockr.org. The second
part is work-in-progress that is jointly done with Alec Wolman (MSR) and Shravan Rayanchu (U. of Wisconsin).
networks (OSNs). Lockr offers three significant privacy benefits to OSN users. First, it separates social networking
content from all other functionality that OSNs provide. This decoupling puts users in control of their own social
information: they decide which OSN providers should store it, which third parties should have access to it, or they can
even choose to manage it themselves. Second, Lockr ensures that digitally signed social relationships needed to access
social data cannot be re-used by an OSN site for unintended purposes. This feature drastically reduces the value of the
social content that users entrust to their OSN providers. Finally, Lockr enables message encryption using a social
relationship key. This key lets two strangers with a common friend verify their relationship without exposing it to
others, a common privacy threat when sharing data in a decentralized scenario.
In the second part, I will present our ongoing work in improving privacy when users run their code in the
infrastructure. This is becoming an increasingly common scenario as cloud computing and mobile computing are becoming
more popular. I will start by discussing these scenarios, after which I will present our current work in minimizing the
attack surface exposed to malicious users and operators in VM-based cloud environments.
Lockr is joint work with Alec Wolman (MSR), Amin Tootoonchian, and Yashar Ganjali (U. of Toronto). For more information
or to download our Lockr implementations for Flickr and for BitTorrent, please visit http://www.lockr.org. The second
part is work-in-progress that is jointly done with Alec Wolman (MSR) and Shravan Rayanchu (U. of Wisconsin).
Contact
Brigitta Hansen
0681 - 9325691
--email hidden
passcode not visible
logged in users only
Uwe Brahm, 02/03/2010 15:24
Carina Schmitt, 12/10/2009 13:14
Brigitta Hansen, 11/30/2009 13:38 -- Created document.
Carina Schmitt, 12/10/2009 13:14
Brigitta Hansen, 11/30/2009 13:38 -- Created document.