Further Adventures in Evidence-Based Security

Subscribers:
344,000
Published on ● Video Link: https://www.youtube.com/watch?v=eChFE0VB5vo



Duration: 1:02:26
1,240 views
5


Over the last decade, our group has undertaken an "evidence-based" approach to measure, analyze, and undermine various kinds of abuse and cybercrime on the Internet. In this talk, I will describe our evidence-based approach and present work on two recent projects. The first is a technique for inferring site compromise from a third-party perspective using honey accounts, and a measurement study that detected 19 site compromises over a year (including an Alexa top-500 site) and our experiences disclosing our findings to these sites. The second is a measurement study exploring the impact of ICANN's New gTLD Program on the domain name ecosystem, classifying the intent of registrations in new TLDs (including speculative and defensive registrations) and impact on old TLDs. I will end briefly summarizing where our group is headed going forward.

This work reflects a long-standing collaborative effort between UCSD and ICSI/Berkeley, with more information available at: http://www.evidencebasedsecurity.org

See more at https://www.microsoft.com/en-us/research/video/adventures-evidence-based-security/




Other Videos By Microsoft Research


2018-04-04Neural Representations for Program Analysis and Synthesis
2018-04-04The Gift of Failure: Fostering Intrinsic Motivation and Resilience in Kids
2018-04-02NeuroSAT: Learning a SAT Solver from Single-Bit Supervision
2018-04-02Using Efficient Oblivious Computation to Keep Data Private and Obfuscate Programs
2018-04-02Neural Models for Information Retrieval
2018-04-01DL Workspace: Running TensorFlow
2018-04-01Designing Gaze Visualizations for Remote Collaboration and Learning
2018-03-29When Psychology Meets Technology with Dr. Daniel McDuff
2018-03-24Monopoly Without a Monopolist: An Economic Analysis of Bitcoin
2018-03-23Enabling Low-power Computation, Sensing, and Communication on Internet-of-Things
2018-03-23Further Adventures in Evidence-Based Security
2018-03-23Recent Results on Learning Filters and Style Transfer
2018-03-23Closing the Loop on Secure System Design
2018-03-22Exploiting Structure Information in Machine Learning
2018-03-22Natural Language Processing: State-of-Art: Observations and Insights
2018-03-22State of Computer Vision
2018-03-22Academic Research Summit 2018 - Welcome & Introductory Session
2018-03-21Generalization Error Bounds for Extreme Multi-class Classification
2018-03-21A Reduction Principle for Generalizing Bona Fide Risk Bounds in Multi-class Setting
2018-03-21Precision-Recall versus Accuracy and the Role of Large Data Sets
2018-03-21Active Positive Semidefinite Matrix Completion With Applications to Large Scale Bandit Problems



Tags:
microsoft research