CS & EE Seminar: Secure and utility-aware data collection with condensed03-01-2020

Speaker: Emre Gürsoy, Georgia Tech

Title:  Secure and utility-aware data collection with condensed local differential privacy

Date/Time: January 3, 2020  /  13.40-14.30

Place: FENS G035

Abstract: Organizations and companies are becoming increasingly interested in collecting customer data and telemetry to make data-driven decisions. While collecting and analyzing customer data is beneficial to improve services and products, customers' privacy poses a major concern. Local Differential Privacy (LDP) has recently emerged as a popular notion for privacy-preserving data collection and has been implemented in several products of companies such as Apple, Google, and Microsoft. However, although existing LDP protocols offer high utility for large customer populations, they perform poorly in scenarios with small populations; and lack perturbation mechanisms that are effective in protecting the content and length of sequential data simultaneously. In this talk, I will introduce the notion of Condensed Local Differential Privacy (CLDP) and a suite of protocols satisfying CLDP for privacy and utility-preserving data collection. The proposed protocols enable the collection of various types of user data in the cybersecurity domain, ranging from ordinal data types in finite metric spaces (malware infection statistics), to non-ordinal items (OS versions and transaction categories), and to sequences of security event flags. Using cybersecurity data and case studies from Symantec, a major cybersecurity vendor, I will show that CLDP protocols are practical for key tasks including malware outbreak detection, OS vulnerability analysis, and inspecting suspicious activities on infected machines.

Bio: M. Emre Gursoy is a PhD student in the School of Computer Science at Georgia Institute of Technology. Prior to Georgia Tech, he received his MS in Computer Science from University of California Los Angeles (UCLA) and his BS in Computer Science and Engineering from Sabanci University. At Sabanci University, he was a recipient of the Dilek Sabanci Scholarship and the Sakip Sabanci Encouragement Scholarship, as well as a merit scholarship for his ranking in the country-wide university entrance exam (OSS). Emre's research interests include data privacy, cybersecurity, adversarial machine learning, location privacy, and big data analytics. He has co-authored over 20 publications in these areas, including several publications in leading IEEE and ACM journals and conferences. He is currently also serving as the Information Director of ACM Transactions on Internet Technology.

Contact: Öznur Taştan