Toward Analysis of Collaborative Cyber Attacks
Exhibit Code: GOL-104
Zone: Computer Zone
Location: Golisano Hall (GOL/070) - Atrium
Time: All Day
In the demo, we will present our research work on network security topic. The work investigates collaborative cyber attacks by exploring the label sequences defined based on spatial patterns extracted from Attack Social Graph. Features called Heavily Attacked Target and Heavy Attacking Conspirator are defined based on ASGs to differentiate hosts in large-scale attacks. Over time, each attack source, which executes multiple attack actions, is associated with a label sequence. A model is trained to describe behavior transitions of the attack sources over time. The model analysis effectively reveals the Internet attack trends and patterns. The experiment shows that the spatial label patterns provide insights on the overall trends and detailed hacking strategies for Internet attacks.