Blockchain-Enabled: Multi-Layered Security Federated Learning Platform for Preserving Data Privacy

Round 1

Reviewer 1 Report

The manuscript entitled "Block Chain-Enabled: Multi-Layered Security Federated Learning Platform for preserving data privacy” is an interesting study and I really enjoyed reading the paper. However, in my opinion, the manuscript has some shortcomings in the text. According to the mentioned items, I recommend minor revision for the manuscript. 

1. Try to polish the writing and English of the manuscript. I found several errors.
2. The authors should provide quantitative results in both the abstract as well as conclusion section.
3. The paper has not suitable structure. The authors should arrange the initial actions to the proposed structure: 1. Introduction, 1.1. statement, and purposes, 1.2. literature review, 2. Methodology 3. result and discussion Conclusion
4. The author should describe the detail of proposed system.
5. The authors did not clarify the effective parameters in this paper. They can mention the details in the abstract and methodology sections.
6. Tables have not been cited in specific order.
7. Please clarify the research gap as well as the innovation of this research at the end of Introduction section.
8. The authors should use more references in 2022 since in the references of this manuscript, there is no paper in these years. Unfortunately, the paper hasn’t had strong literature review. They can study and address the following paper:

https://doi.org/10.3390/s22020450

https://doi.org/10.1016/j.trc.2020.01.002

https://doi.org/10.1016/j.scs.2020.102220

 

9. The flowchart of study should be revised according to the study goal.
10. The conclusion section should represent the results on a case-by-case basis by bullets.
11. Why haven’t authors used other DL and Machine learning methods to model?
12. It is better author compare their result with previous studies in order to present state of art
13. The authors should provide future research at the end of the conclusion section. They can recommend other machine learning techniques to be incorporated into the proposed approaches to obtain more accurate results:

https://doi.org/10.1155/2021/9048808

https://doi.org/10.1007/s11709-021-0785-x

https://doi.org/10.1155/2021/9984840

https://doi.org/10.1155/2021/9974219

https://ieeexplore.ieee.org/abstract/document/9420107

https://ieeexplore.ieee.org/abstract/document/9317334

https://doi.org/10.1016/j.conbuildmat.2020.120509

 

 

Comments for author File: Comments.pdf

Author Response

Respected Reviewer, 

1. Try to polish the writing and English of the manuscript. I found several errors.

Response: I try to resolve this issue 

1. The authors should provide quantitative results in both the abstract as well as conclusion section.

Response: Task completed as per Instructions.

1. The paper has not suitable structure. The authors should arrange the initial actions to the proposed structure: 1. Introduction, 1.1. statement, and purposes, 1.2. literature review, 2. Methodology 3. result and discussion Conclusion.

Response: Document aligned 

The author should describe the detail of proposed system.

Response: I describe More detail as well performed more simulations.

The authors did not clarify the effective parameters in this paper. They can mention the details in the abstract and methodology sections.

Response: Comment addressed in Document.

Please clarify the research gap as well as the innovation of this research at the end of Introduction section.

Response: Comment addressed in Document.

The authors should use more references in 2022 since in the references of this manuscript, there is no paper in these years. Unfortunately, the paper hasn’t had strong literature review. They can study and address the following paper:

Response: Comment addressed and wrote some specific points in that section. 

The flowchart of study should be revised according to the study goal.

Response: Flow chart revised. 

The conclusion section should represent the results on a case-by-case basis by bullets.

Response : Aligned Case by case Results.

It is better author compare their result with previous studies in order to present state of art.

Response: different modelComparison  and simulations performed .

Why haven’t authors used other DL and Machine learning methods to model?

Response: Added DL and machine learning Methods.

The authors should provide future research at the end of the conclusion section.

Response: Comment Addressed in Document. 

 

 

 

 

Author Response File: Author Response.pdf

Reviewer 2 Report

The article discusses about multi-layer FL platform for upgrading data privacy. Few major issues are found in the article that minimizes the quality of the paper as below.

1. Content wise novelty of the work is very minimal. Section 2 and 3 have no significant role in quality literature of this work. Section 2 and 3 must be revised in order to cope up with the key objective of the work.
2.  No clear motivation neither objectives are mentioned that make the readability of the paper very poor.
3. System model section is not discussed in detail at all. Figure 1 and 2 are poorly presented with very less significance to the work. They must be improved.
4.  Algorithm 1 is not elaborated in detail. What is the complexity of the algorithm 1?
5. Proposed blockchain model as mentioned in section 4 is not adequate. Privacy-preserving flow chart in Figure 3 is not described well to align with the privacy preservation policy of this work.
6. The simulation results must need comparison with similar related works. No such mention is present in current form.
7. Table 4 and 5 must be clearly discussed with key relationship with the work.
8. How poisoning attack is resolved in this work is not clear! Conclusion part doesn't show how this work can further improved.
9. More recent papers must be include din the reference section.

 

Overall presentation of the paper is not of good quality and must be addressed with detailed language and technicalities. More emphasis must be given while drawing diagrams and presenting algorithms. Few mathematical statements are also included in the paper, however no correlation discussion is clearly made.

Author Response

Please check the response attached.

Author Response File: Author Response.pdf

Reviewer 3 Report

The topic is new, interesting, and highly targeted by researchers.

The previous studies and history of this work are missing from the paper.

The proposed algorithm could be presented more clearly and systematically using Pseudocode Instead of flowcharts. 

An extensive set of experiments are required to prove the effectiveness and importance of the proposed work.

The written paper needs proofreading.

 

Author Response

The previous studies and history of this work are missing from the paper.

Response: Specific History of Work added in document.

The proposed algorithm could be presented more clearly and systematically using Pseudocode Instead of flowcharts. 

Response : Polished Flow chart and provided more Pseudo code and simulations.

An extensive set of experiments are required to prove the effectiveness and importance of the proposed work.

Response: Added a set of experiment to prove effectiveness.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The version is improved.

Author Response

Improvements has been made in article. 

Reviewer 3 Report

The paper has been improved.

However, the introduced pseudo-codes can be enhanced.

the conclusion should be re-written

the flowchart should be changed into meaningful pseudo code.

The introduction section should be divided into paragraphs.

 

 

Author Response

 

 

1. the conclusion should be re-written.

onclusion Upgraded.

2. the flowchart should be changed into meaningful pseudo code.

Flow chart and Pseudo code both Entere.

3. The introduction section should be divided into paragraphs.

Done