Abstract
This paper tackles current challenges in network security analysis by proposing an innovative information gain‐based feature selection algorithm and leveraging visualization techniques to develop a network security log data visualization system. The system’s key functions include raw data collection for firewall logs and intrusion detection logs, data preprocessing, database management, data manipulation, data logic processing, and data visualization. Through statistical analysis of log data and the construction of visualization models, the system presents analysis results in diverse graphical formats while offering interactive capabilities. Seamlessly integrating data generation, processing, analysis, and display processes, the system demonstrates high accuracy, precision, recall, F1 score, and real‐time performance metrics, reaching 98.3%, 92.1%, 97.5%, 98.1%, and 91.2%, respectively, in experimental evaluations. The proposed method significantly enhances real‐time prediction capabilities of network security status and monitoring efficiency of network devices, providing a robust security assurance tool.
Publisher
Institution of Engineering and Technology (IET)