UQ NIDS (ML-based NIDS NetFlow Datasets)

The datasets on this page are designed for machine learning-based Network Intrusion Detection Systems (NIDS) and are organised into the following high-level collections:

NetFlow V3 Datasets: This collection consists of four datasets in NetFlow format. This is the only version incorporating temporal features. These datasets build on the 43 features of V2 by adding 10 temporal NetFlow features, totaling 53 features. They are provided in CSV format.
NetFlow V2 Datasets: Comprising five datasets in NetFlow format, this collection extends the V1 datasets with 43 enhanced NetFlow features. They are available in CSV format.
NetFlow V1 Datasets: This collection includes five datasets converted from four diverse formats into NetFlow format, featuring 12 basic NetFlow features. They are provided in CSV format.
CICFlowMeter Datasets: Consisting of five datasets, this collection is generated using the CICFlowMeter tool by converting three datasets from various formats into CICFlowMeter format, with 80 features. They are available in CSV format.

For full details and the recommended citation for these datasets, please refer to the corresponding links in the tabs below.