Abstract
This study, explores the potential of browser log analysis for constructing comprehensive digital profiles of users. A dataset of 138,105 records collected from a university network through the Fortigate logging system serves as the basis for an extensive examination of web traffic activity. The methodology integrates descriptive statistical analyses, time-of-day evaluations and advanced machine learning techniques, including K-means and DBSCAN clustering for user segmentation, along with Isolation Forest and Local Outlier Factor for anomaly detection. Quantitative metrics—such as aggregate sent and received data volumes by date and hour, session counts, application usage by category and risk assessments—were computed and visualized to reveal distinct behavioral patterns. The results indicate significant variations in user activity, highlighting peak usage periods and categorizing users into groups that represent routine, collaborative and high-risk behaviors. These findings demonstrate that browser log analysis is an effective tool for digital profiling, with practical applications in personalized service delivery and network security enhancement. The study also emphasizes the importance of ethical data handling and privacy-preserving measures. Limitations related to the scope of data and classification accuracy are discussed and future research directions are proposed to further refine profiling techniques by incorporating additional data sources.