CICIDS2017 and UNBSW-NB15

The UNSW-NB15 Dataset is a compilation of raw network data packets crafted by the University of South Wales. This dataset is designed to create a blend of modern normal network activity and synthetic contemporary attack behavior. It encompasses nine types of attacks, including Fuzzers, Analysis, Backdoors, DoS, Exploits, Generic, Reconnaissance, Shellcode, and Worms, resulting in a total of 10 classes of traffic and 49 features.

The CICIDS2017 dataset was curated by the Canadian Institute for Cybersecurity (CIC) using the B-Profile system, which simulates a network environment with a continuous variety of traffic attacks over a five-day period. This dataset serves as a popular training resource, containing both benign and up-to-date common attacks, resembling real-world data (PCAPs). It encompasses a total of 15 labels, including 14 network attacks and normal traffic, such as Botnet, Web Attack Brute Force, DoS, DDoS, Infiltration, Heartbleed, Bot, PortScan, and other common attacks based on the McAfee report of 2016.Furthermore, it includes 15 labels for traffic based on the results of network traffic analysis performed by CICFlowMeter. The analysis incorporates labeled flows derived from timestamps, source and destination IP addresses, source and destination ports, protocols, and attacks, all presented in CSV files