2022 IEEE Conference on Communications and Network Security (CNS) 2022
DOI: 10.1109/cns56114.2022.9947235
|View full text |Cite
|
Sign up to set email alerts
|

Error Prevalence in NIDS datasets: A Case Study on CIC-IDS-2017 and CSE-CIC-IDS-2018

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

1
3
0

Year Published

2023
2023
2024
2024

Publication Types

Select...
8
1

Relationship

0
9

Authors

Journals

citations
Cited by 27 publications
(7 citation statements)
references
References 31 publications
1
3
0
Order By: Relevance
“…In fact, we split and labeled the PCAP data based on the attack paths, attack times, and other information from official documents, rather than the CSV files provided. According to our analysis, we found some issues that are consistent with the findings of Liu (2022) [41]. For example, in CSE-CIC-IDS2018, the execution description of BruteForce FTP can not be matched with the PCAP data, so we removed it from the attack list.…”
Section: Dataset and Experimental Setupsupporting
confidence: 78%
See 1 more Smart Citation
“…In fact, we split and labeled the PCAP data based on the attack paths, attack times, and other information from official documents, rather than the CSV files provided. According to our analysis, we found some issues that are consistent with the findings of Liu (2022) [41]. For example, in CSE-CIC-IDS2018, the execution description of BruteForce FTP can not be matched with the PCAP data, so we removed it from the attack list.…”
Section: Dataset and Experimental Setupsupporting
confidence: 78%
“…Liu et al analyzed the errors [41] present in both datasets. They analyzed the inconsistency of the dataset labels with the actual attack behavior and provided a modified dataset.…”
Section: Dataset and Experimental Setupmentioning
confidence: 99%
“…Kin et al [11] developed a Convolutional Neural Network based model for the detection of DDoS attacks using the KDD [12] and CSE-CIC-IDS 2018 [13] datasets. Focusing on improving intrusion detection systems, the study addressed the challenge of distinguishing DoS attacks, including advanced types, from benign traffic.…”
Section: Related Work a Machine Learning In Ddos Detectionmentioning
confidence: 99%
“…However, this approach proves to be inefficient as it requires filtering the output CSV files N times, with N being the line number of the public tabular data. Moreover, it is worth noting that some errors have been identified in certain public tabular datasets [32,33] recently, further complicating the efficiency and reliability of this method.…”
Section: Flow Construction and Labeling Modulementioning
confidence: 99%