Statistical and Multivariate Analysis of the IoT-23 Dataset: A Comprehensive Approach to Network Traffic Pattern Discovery
Round 1
Reviewer 1 Report
This research presents a comprehensive analysis of the IoT-23 dataset, a labeled IoT network traffic dataset containing both benign and malicious records. The findings can contribute to the understanding of IoT network behavior and provide a foundation for developing robust intrusion detection systems. This work establishes a methodological framework for analyzing IoT network traffic datasets. Generally, the paper is well-written and makes some scientific contributions. However, it has some few flaws that must be addressed to enhance its relevance and quality.
1) The focus of this paper is Network Traffic Pattern Discovery. However, the abstract has not described this research area (including the problem domain).
2) To prevent reader confusion, all acronyms must be written in full the fist time they appear within text.
3) In the abstract, you claim that statistical measures including mean, variance, skewness, and kurtosis provided insights into the dataset's non-normal distributions.
-Support the above claim by some numerical data.
4) In the Introduction section, you write as follows: "the research community is actively working to address the challenges faced by this thriving sector. To support their efforts, several IoT datasets have been released, including CICIoT2023, EdgeIIoT, IOT-23, and TON_IoT, which are available in the public domain."
i) Which specific 'challenges faced by this thriving sector' are you referring to?
ii) Give details of the mentioned datasets. What are the contents and rationale for these datasets?
5) The benefits of statistical data analysis given in Section 1 are well-known and hence no need to list them here.
6) The Introduction section is too brief and fails to vividly bring out the research area and problem domain.
7) The essence of the Related Works in Section 2 is to bring out some research gaps. Therefore, you need to describe some of the gaps of each of the presented works. Towards the end of this section, add a paragraph to summarize the identified research gaps. Thereafter, explain how you managed to bridge these gaps.
8) The citation in the first paragraph of section 3 is not well formatted (check the journal guidelines). What do you imply by '(Garcia et al., 2020) [3.14]'?
9) In the Descriptive Measures of Section 5, you have presented some graphs (Figure 1, Figure 2, Figure 3, Figure 4 etc).
i)You have not specified the particular dataset(s) used to draw these graphs
ii) What is the need for these graphs since they do not pertain to any of the datasets?
iii) Revise all figures to specify the dataset deployed.
10) Based on the presented results:
i) What is the main deliverable of this work?
ii) How are the findings related to security/privacy?
11) In Table 8, you provide some comparison with existing research.
i) Based on Table 8, you used IoT-23 dataset. What was the rationale behind this section?
ii) What uniqueness of your approach can the readers pick from Table 8?
12) Revise the conclusion section to describe some of the limitations of this study. Afterwards, explain how these limitations can be addressed. What are some of the feasible future research scopes in this domain?
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The study employs descriptive statistics and multivariate analysis for understanding data distributions of the IoT-23 dataset, which contains labeled IoT network traffic data (both benign and malicious). This work tries to uncover traffic patterns and relationships among features, which is meaningful in practice. The reviewer thinks there will be many readers.
There have been similar studies on other datasets and the techniques using in this study are not original. However, t for the IoT-23 dataset, it seems that no comparable work has been conducted yet.
Generally, this manuscript is written well and easy to follow. But, there are still some places to be improves.
1) It would be better if the authors explained more carefully how useful their analysis results are in practice, for example in feature selection.
2) The size/postion of Table 5 should be adjusted.
3) How to find patterns related to time-series should also be discussed.
4) In Reference, numbering should be revised.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
The study applies different methods of statistical analysis, in a combination that has not been used in other studies, and a comparison is made with existing solutions. The focus of the study is the IoT-23 database and the analysis of data related to attacks in it.
It is not clear why the sources are numbered as 3.1, 3.2, 3.3, etc. The template requirements are not like that. It is suggested to review the template again.
The template does not indicate that et al citations or citations using authors names can be used. It is recommended to check the template again.
Tables are not formatted according to the requirements of the template.
Figure 1 should be formatтed according requirements of the template.
The equations are not formatted according to template, they are into cells, please check the template again.
The closing parenthesis of the value 0.313253 on page 14 line 406 is missing.
It is proposed to check the requirements for the template for numbering references.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
This research presents a comprehensive analysis of the IoT-23 dataset, a labeled IoT network traffic dataset containing both benign and malicious records. The findings can contribute to the understanding of IoT network behavior and provide a foundation for developing robust intrusion detection systems. This work establishes a methodological framework for analyzing IoT network traffic datasets. Generally, the paper is well-written and makes some scientific contributions.
Thank you for effectively responding to all the previous comments.
