© CyberShield : AI powered broken access control detection and mitigation
Journal of Information & Optimization Sciences · 2026 · 47 (5-B) · 1907-1917
DOI: 10.47974/jios-2283 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Information & Optimization Sciences, 2026. Verfasst von Aashmit Mckenzie und 6 weiteren Autor:innen.
Abstract (Originalsprache)
Broken Access Control (BAC) is identified as the most dangerous and common vulnerability in the modern web systems. BAC can cause severe problems, such as data leaks, privilege escalations and unauthorized system manipulations. To address this issue, we propose to manually construct an extensive and well-featured dataset that can represent realistic traffic-based access control behaviors and behaviors of BAC attacks. This dataset serves as a preliminary resource for developing and evaluating machine learning algorithms capable of real-time BAC attack detection. The study has employed this dataset to develop various machine learning techniques such as Logistic Regression, Random Forest, Gradient Boosting, and XGBoost. Double is the contribution of the proposed work, namely the new BAC dataset that serves as a fundamental requirement for research communities, and a machine learning-based detection framework that is dedicated for real-time attacks detection. The Random Forest classifier achieved the highest overall accuracy (90.5%) and F1- score (0.8627), while XGBoost had the greatest AUC value of 0.9556. The system can also
Zitieren
APA
Mckenzie, A., Doshi, M., Deoghare, G., Bidwe, R., Kothari, S., Bagane, P., & Goyal, A. (2026). © CyberShield : AI powered broken access control detection and mitigation. Journal of Information & Optimization Sciences, 47(5-B), 1907-1917. https://doi.org/10.47974/jios-2283
Vancouver
Mckenzie A, Doshi M, Deoghare G, Bidwe R, Kothari S, Bagane P, et al. © CyberShield : AI powered broken access control detection and mitigation. Journal of Information & Optimization Sciences. 2026;47(5-B):1907-1917. doi:10.47974/jios-2283
BibTeX
@article{Mckenzie2026,
author = {Mckenzie, Aashmit and Doshi, Malay and Deoghare, Gauri and Bidwe, Ranjeet and Kothari, Sonali and Bagane, Pooja and Goyal, Ankur},
title = {© CyberShield : AI powered broken access control detection and mitigation},
journal = {Journal of Information & Optimization Sciences},
year = {2026},
volume = {47},
number = {5-B},
pages = {1907--1917},
doi = {10.47974/jios-2283}
}
EUVIMED