Intrusion Detection Systems (IDS) plays an important role in web mining to detect and prevent unauthorized access, attacks, and anomalies in web-based systems. Monitoring unauthorized or malicious activity on networks or systems is the purpose of IDS. Building an effective intrusion detection system for web mining is a complex task that requires expertise in web security, data analysis, and machine learning. By implementing data mining in a cyber-attack intrusion detection system, organizations can sense and react to threats more rapidly and effectively, reducing the hazard of harm and loss. Data mining can also provide insights into the nature and characteristics of cyber-attacks, which can inform the development of more effective security measures. In this paper, log files are collected and preprocessed using Kalman filter with the combination of Correlation-based Feature Selection (CFS). The experimental findings demonstrate that such pre-processing and CFS combination is effective are applied with different classification algorithms in machine learning models using KDD’99 dataset.