Panel: Big Data Sharing for Security: Is “Big Data” Worth It?

Moderator: Sven Dietrich, John Jay College of Criminal Justice & The Graduate Center, City University of New York, http://jjcweb.jjay.cuny.edu/sdietrich/

Panelists:
Jean Camp, Indiana University at Bloomington
Michael Collins, RedJack
David Dittrich, University of Washington

Abstract:
As the state-of-the-art big data collection efforts bring together information from multiple security and non-security data sources, such as the network, the desktops and servers, the embedded devices in our public spaces and private homes, we hope to derive newfound insights for normal and malicious behavior using machine-learning and data mining techniques. Moreover, in order to better understand the Internet and their users, we need to ensure reproducibility of these results and we may (must?) share the raw or processed data sets for independent verification and validation. With that comes a dilemma: how do we balance the technical ability to share vs. the intended values, and the realized values of sharing? While there are now frameworks in existence to describe sharing of data sets, as well as ethical guidelines in the form of the Menlo Report, we still face challenges ahead, especially when it comes to their technical implementation and the recognition of the social impact.

The panel discusses such ongoing challenges arising from several contexts, such as ethical data sharing with human subjects, network data sharing from controversial/criminal cases, and large-scale (Internet-scale) data collection and the related anonymity and privacy concerns. The panelists are from multiple disciplines, backgrounds, and affiliation to represent their respective prioritized agenda for big data for security.

SLIDES