With the rise of digital transformation, web security has become a critical concern for organizations, governments, and individuals. This study explores the role of data science in enhancing web security by leveraging machine learning algorithms and advanced analytics to predict and identify potential attacks in real-time. The main objective is to demonstrate how data-driven techniques, including predictive analytics, anomaly detection, and behavioral analysis, can be integrated into existing security frameworks to reduce vulnerabilities and strengthen defenses against cyber threats. The research gap addressed by this study lies in the insufficient application of comprehensive, data-driven methodologies for threat detection and classification in web security. The problem gap is the absence of integrated frameworks that combine feature engineering, classification models, and anomaly detection for both known and unknown threats. This study bridges these gaps by employing a structured dataset of web interactions to model, detect, and predict security threats using advanced data science techniques. Using a dataset of simulated web traffic and previous attack records, this research applies data preprocessing, feature engineering, and machine learning classification models, such as decision trees and random forests, to predict threat levels and identify anomalies. Results show that machine learning models can effectively classify threat levels, with a threat classification accuracy of 80 percent. This study contributes to the field by demonstrating how data science can improve web security practices, offering a proactive approach to detecting and mitigating cyber-attacks.
