Abstract

Cyber threats have become increasingly sophisticated, with ransomware attacks posing a particularly disruptive risk by encrypting critical data and demanding ransoms. Traditional detection methods, which primarily rely on signature or heuristic-based models, often fall short in identifying advanced ransomware strains, especially those employing obfuscation and polymorphic techniques to evade detection. The Adaptive Behavior Profiling System (ABPS) presents a novel solution, leveraging dynamic behavior modeling alongside network pattern profiling to enhance detection efficacy. Through a multi-layered approach, ABPS achieves high accuracy, sensitivity, and specificity by identifying distinct behavioral and networkbased indicators of ransomware activity. Comparative evaluation against established techniques highlights ABPS's superiority in reducing false positives and maintaining low latency, ensuring its practical applicability in real-time cybersecurity environments. Results show that ABPS consistently outperforms traditional detection models, demonstrating a significant advancement in ransomware detection through its adaptive, resource-efficient profiling strategy. By balancing robust detection capabilities with operational efficiency, ABPS contributes a versatile and scalable solution to the ongoing challenges of ransomware detection.