Location
Hilton Hawaiian Village, Honolulu, Hawaii
Event Website
https://hicss.hawaii.edu/
Start Date
3-1-2024 12:00 AM
End Date
6-1-2024 12:00 AM
Description
Becoming a reverse engineer (RE) requires rigorous training and understanding of program structure and functionality, and experts develop heuristic strategies and intuitions from real-world experiences. This paper attempts to capture REs’ strategies and intuitions within a predictive cognitive model and demonstrate the feasibility of assisting novice REs using an intelligent recommender called CAVA (Cognitive Aid for Vulnerability Analysis). CAVA leverages physiological sensors to assess a novice’s cognitive states and provides real-time visual hints when the novice’s attention and engagement diminish. We instrumented Ghidra and conducted pilot experiments with REs. Open-loop experiments with 9 REs confirmed the feasibility of identifying novices from experts using physiological signals, and a pilot closed-loop experiment tested the feasibility of providing visual recommendations to a novice. Despite challenges in recruiting REs, our progress suggests that CAVA is a promising approach to improve novice performance and our understanding of experts’ behavior when performing complex real-world reverse engineering tasks.
Recommended Citation
Kim, Evelyn; Fugate, Sunny; Lebiere, Christian; Barbieux, Aidan; Buch, Jonathan; Cho, Jaehoon; Cranford, Edward; Divita, Joseph; Johnson, Jeremy; Levy, Mia; Maldonado, Froylan; Marsh, Brianna; and Morrison, Donald, "CAVA: Cognitive Aid for Vulnerability Analysis" (2024). Hawaii International Conference on System Sciences 2024 (HICSS-57). 2.
https://aisel.aisnet.org/hicss-57/st/digital_forensics/2
CAVA: Cognitive Aid for Vulnerability Analysis
Hilton Hawaiian Village, Honolulu, Hawaii
Becoming a reverse engineer (RE) requires rigorous training and understanding of program structure and functionality, and experts develop heuristic strategies and intuitions from real-world experiences. This paper attempts to capture REs’ strategies and intuitions within a predictive cognitive model and demonstrate the feasibility of assisting novice REs using an intelligent recommender called CAVA (Cognitive Aid for Vulnerability Analysis). CAVA leverages physiological sensors to assess a novice’s cognitive states and provides real-time visual hints when the novice’s attention and engagement diminish. We instrumented Ghidra and conducted pilot experiments with REs. Open-loop experiments with 9 REs confirmed the feasibility of identifying novices from experts using physiological signals, and a pilot closed-loop experiment tested the feasibility of providing visual recommendations to a novice. Despite challenges in recruiting REs, our progress suggests that CAVA is a promising approach to improve novice performance and our understanding of experts’ behavior when performing complex real-world reverse engineering tasks.
https://aisel.aisnet.org/hicss-57/st/digital_forensics/2