Data Confidentiality Workshop
Home Workshop Agenda Participants Travel Information

 

Contact

 


WORKSHOP ON DATA CONFIDENTIALITY

September 6-7, 2007 in Arlington, VA

White Paper & Bio


The availability of realistic network data plays a significant role in fostering collaboration and ensuring technical leadership in network security research. Unfortunately, a host of technical, legal, policy, and privacy issues limit the ability of operators to produce datasets for information security testing. One recent solution to providing low-risk, high-value data is that of trace anonymization---a process of sanitizing data before release so that information of concern cannot be extracted.

Network trace anonymization itself is preceded by decades of established practice in the medical community to anonymize patient records that are used in medical research. Network traces embody a similar tension as that which faces the medical community: They are necessary for a wide range of research in the trends of and remedies to network ``diseases'' (e.g., malware). They also embody significant information about the networks and individuals represented in the trace, some of which could be similarly damaging if revealed (e.g, security posture and procedures of an organization).

While privacy for both medical records and network traffic records is a problem of statistical inference control, each domain offers to an adversary different methods to gain information that might be used to deanonymize a dataset and requires that different types of information be preserved in the dataset. Indeed, the current uncertainties about network trace anonymization, from both technical and policy perspectives, leave the research community in a vulnerable position. Even as it marches forward with invaluable data collection efforts, it does so with little understanding of the privacy violations that may result and with no tools or techniques for evaluating the privacy risks of disclosing a dataset. Addressing this challenge is essential, to ensure the continued, yet responsible, availability of network trace data to support security research. It is this challenge that we would like to see the research community address in the near term.

Fabian Monrose

Johns Hopkins

 

Biographical Data

Fabian Monrose is an Associate Professor in the Computer Science department at Johns Hopkins. He holds a joint appointment with the Hopkins Information Security Institute. Prior to joining Hopkins he was a member of the Secure Systems Group at Bell Labs, Lucent Technologies. His research interests include computer and network security, applied cryptography, data anonymization and privacy preserving techniques, biometrics, network traffic classification, to name a few.

See http://cs.jhu.edu/~fabian for more information.