Soteria: An Approach for Detecting Multi-Institution Attacks
Saif Zabarah, Omar Naman, Mohammad A. Salahuddin, Raouf Boutaba, Samer Al-Kiswany · 2023
We present Soteria, a data processing pipeline for detecting multi-institution attacks. Soteria uses a set of Machine Learning techniques to detect future attacks, predict their future targets, and ranks attacks based on their predicted severity. Our evaluation with real data from Canada wide academic institution networks shows that Soteria can predict future attacks with 95% recall rate, predict the next targets of an attack with 97% recall rate, and detect attacks in the first 20% of their life span. Soteria is deployed in production and is in use by tens of Canadian academic institutions that are part of the CANARIE IDS project.