I graduated in May and am planning my next steps.
I was a member of Systems Research at Harvard (Syrah). I started out working on the Provenance Aware Storage Systems (PASS) project. My focus was on understanding what provenance can and cannot do and how to secure the provenance we collected. In the process of working on securing provenance, I realized my approach and solution are not limited to provenance as they apply to annotated graphs more generally.
Harvard is unique in that we students can and do put together teams of Professors to advance us and our research. I am especially indebted to Professors Stephen Chong and Greg Morrisett for advising me on the program language aspects of my research, Professor Jim Waldo for help ensuring my work was well founded, and to Professor Latanya Sweeney for her help relating to security and privacy.
I prefer taking the difficult path as I learn more that way. My research focuses on limiting disclosure of data represented in annotated graphs. I have proven that graphs provide richer semantics than earlier data models and introduce new privacy challenges. Graphs require new and richer: policies, security models, and enforcement mechanisms. Society is plowing forward, but the technical and legal communities are far behind.
Examples of annotated graphs include: social networks, medical records, financial data, and national intelligence. The key difference between annotated graphs and traditional (e.g. file or database) models is that graphs lacks a single point of control and inferences are often global. For example, to hide that two people are friends, we cannot limit the controls to one of the friends as both participants can reveal the friendship. Furthermore, a large overlap in circles of friends suggests the existence of the friendship.
This work is at the nexus of: systems, security and privacy, and programming language theory. I believe in: disproving existing approaches, proving mine works, building it, and showing it is practically useful. Anything short of achieving all four is a work in progress.
More broadly, I am interested in distributed systems -- getting computers to “play nice” together, especially making sure they continue to play nice even when life is not perfect.
I am active in behind the scenes roles in government on campus. I focus on the areas where we all agree, because there is plenty to do there. My efforts span the area, school and university level. At the area level, I am instrumental in starting and running cs-community as an informal community primarily targeting graduate students and CS undergrad concentrators. This includes putting on events and soliciting discussion on concerns. SEAS is the newest school at Harvard and still lacks a student governance structure both at the school and area level. I am trying to encourage the creation of organizations like cs-community in SEAS' other areas. At the SEAS level, I have been involved in the Graduate Student Life Council (GSLC). Currently, a team of SEAS students is working to create both a student government organization and a social organization. I have also represented Computer Science and/or SEAS at both the Graduate School of Arts and Sciences (GSAS) Graduate Student Council (GSC) as well as at the University level at the Harvard Graduate Council (HGC).
An entire cast of supporting Professors deserve mention, including: Michael Rabin for teaching me the power of thinking probabilistically, Gerome Miklau for providing informal advice when it mattered, Jim Waldo for his wisdom on both distributed systems and anchoring ones work on philosophical bedrock, Harry Lewis for solving a proof I worked on for months, Ronald Rivest for help on that same proof.
I am far more indebted to my fellow graduate students, who have taught me far more. First among these is Avi Shinnar a true friend, who taught me volumes both academically and otherwise, Gregory Malecha and Ryan Wisnesky both friends who have selflessly helped me academically and improved the sense of community, David Darais an energetic friend and teacher, Bo Waggoner a running friend and thinker.
David Holland deserves special mention. He has steadfastly supported me throughout the years both academically and personally. Easily one of the best programmers I know, his advising skills are his greatest asset.
Lots more to add...
|Uri Braun Last modified: Wed Feb 10 23:21:18 Eastern Standard Time 2010|