A Detailed Analysis of Benchmark Datasets for Network Intrusion Detection System

The enormous increase in the use of Internet daily life has provided an opportunity for intruder attempt to compromise security principles availability, confidentiality, and integrity. As a result, organizations are working level by using attack detection techniques such as Network Intrusion Detection System (NIDS), which monitors analyzes network flow attacks detection. There lot researches proposed develop NIDS depend on dataset evaluation. Datasets allow evaluating ability detecting intrusion behavior. This paper introduces detailed analysis benchmark recent datasets NIDS. Specifically, we describe eight well-known that include: KDD99, NSL-KDD, KYOTO 2006+, ISCX2012, UNSW-NB 15, CIDDS-001, CICIDS2017, CSE-CIC-IDS2018. For each dataset, provide its instances, features, classes, nature features. main objective this is offer overviews available what comprised of. Furthermore, some recommendations were made network-based datasets.