The escalating cybersecurity threats have intensified the need for resilient frameworks capable of accurately detecting and classifying sophisticated ransomware. Traditional approaches often rely on static signatures and heuristic methods, which can struggle against adaptive ransomware variants that employ evasion techniques. Addressing these limitations, the Behavioral Attack Signatures Evaluation (BASE) framework introduces an adaptive, behavior-driven model that utilizes advanced machine learning to recognize ransomware through its distinct behavioral characteristics. Through dynamic analysis of high-impact features, including file access patterns, system call frequency, and network interactions, BASE constructs robust behavioral signatures that differentiate ransomware from benign processes. Experimental results reveal that BASE not only achieves superior detection accuracy and lower false-positive rates compared to signature-based and heuristic models but also demonstrates strong adaptability to new ransomware variants. Additionally, BASE exhibits efficient scalability, maintaining rapid processing times even under high network loads and varying operational conditions. With its capacity to integrate seamlessly within diverse cybersecurity infrastructures, the BASE framework represents a substantial advancement in ransomware detection and classification, providing a powerful tool for organizations in need of proactive and reliable defense mechanisms.