Predicting Impending Exposure to Malicious Content from User Behavior
Association for Computing Machinery (ACM) via YouTube
Save 40% on 3 months of Coursera Plus
AI Engineer - Learn how to integrate AI into software applications
Overview
Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore a groundbreaking approach to computer security in this 23-minute ACM conference talk on predicting and preventing exposure to malicious content through user behavior analysis. Learn how this innovative system enables proactive defenses at the single browsing session level, moving beyond traditional reactive security measures and long-term prediction models. Discover the methodology behind processing user data into sessions, understanding behavioral differences between users, and implementing exposure prediction techniques. Examine the results of this approach, including its effectiveness in mitigating the base-rate effect. Gain insights into the future of cybersecurity and how this research contributes to creating more robust, anticipatory defense mechanisms for individual users.
Syllabus
Intro
Traditional defenses are reactive
Proactive defenses work over long periods
Processing data into sessions
Window of exposure
Behavioral differences between users (2/3)
Exposure prediction: Methodology (1/2)
Exposure prediction: Results (1/2)
Base-rate effect (2/2)
Wrap up
Taught by
Association for Computing Machinery (ACM)