Cybersecurity Architecture That Protects Shared Employees Using Data-Driven Insights

Advances in remote work have brought new challenges to network security, especially when it comes to protecting user credentials. Many factors, such as lack of password management policies, lack of security awareness, and protection of sensitive information on untrusted networks, can lead to this vulnerability. Attackers often exploit these weaknesses to gain unauthorized access to network resources by capturing user credentials from untrusted URLs. This paper introduces a machine learning-based framework to enhance network security by evaluating user credentials and enforcing password management policies. Our method uses various classification methods, including Deterministic Decision (DT), K-Nearest Neighbor (KNN), Random Forest (RDF), and AdaBoost (ADAB), to match legitimate (trusted high) and malicious (untrusted) URLs. Measure the performance of your solutions using metrics such as precision, recall, and F1 score.