NGEN Firewall Security Augmentation using Brooks-Iyengar and Random Forest Classfier method: by Predicting Cyber Threats from: Darkweb/Deepweb Data
Latesh Kumar KJ, H. U. Leena · INTERNATIONAL JOURNAL OF NEXT-GENERATION COMPUTING · 2020
there are many caveats persist in current Unified Threat Management Firewalls (UTM) device for assessing vulnerabilities in software and hardware. The techniques used to exploit threats in UTM appliances are less dynamic non-predictive and hence cybersecurity is still fickle. In this paper we propose a technique to integrate Brooks-Iyengar Fault algorithm and Random Forest Classifier model to analyze the dark web and deep web network analysis using machine learning methods with UTM devices to envision the exploitability of vulnerabilities. Our technique achieves this aim by analyzing vulnerability data from UTM logs, Microsoft and Redhat attack signatures, National Vulnerability Database and features established by user’s association with deepweb and darkweb (d2web) sites. We carry out a real-time experiment on a honey trap case study by involving real-time cyber criminal’s activity and vulnerability data to mitigate cyber risks. The results published are evaluated using F1 score and IPS and IDS improved by 16% while maintaining the performance and precision. We consider this result because many exploit cases recorded and documented of various vulnerabilities with their score are indicative of their ability in exposing the threat and impact, the prediction score by 94.3% shows the actual and subsequent threat analysis results with private cloud and elite firewall policy service.