FG-ENTAD

This repository contains a network traffic dataset derived from real firewall logs collected from a FortiGate Next-Generation Firewall deployed in an institutional network environment. The dataset was constructed to support research on anomaly detection in network traffic using data-centric machine learning approaches.

The dataset is associated with the research article “A Data-Centric Network Traffic Dataset for Anomaly Detection: Construction, Reproducible Pipeline, and Technical Validation.” It follows a reproducible data engineering pipeline including preprocessing, feature engineering, unsupervised pseudo-labeling using EFMS-KMeans clustering, and dataset preparation for sequential deep learning models such as CNN–GRU.

The repository is organized into several stages reflecting the dataset generation process: raw firewall logs, processed traffic data, pseudo-labeled clustering outputs, labeled datasets prepared for machine learning experiments, and metadata files describing dataset attributes and feature definitions.

This dataset is intended to support research in cybersecurity, intrusion detection systems (IDS), anomaly detection, and machine learning-based network traffic analysis.