Next Article in Journal
A Review of Selected Solutions on the Evaluation of Coal-Rock Cutting Performances of Shearer Picks under Complex Geological Conditions
Next Article in Special Issue
A Hybrid CNN and RNN Variant Model for Music Classification
Previous Article in Journal
Influence Analysis of Traveling Wave Effect on Rail Interaction of Long-Span Suspension Bridge
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Applied Sciences
Volume 12
Issue 23
10.3390/app122312367
Review Report
applsci-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
Article
Peer-Review Record

Malicious URL Detection Model Based on Bidirectional Gated Recurrent Unit and Attention Mechanism

Appl. Sci. 2022, 12(23), 12367; https://doi.org/10.3390/app122312367
by Tiefeng Wu, Miao Wang *, Yunfang Xi and Zhichao Zhao
Reviewer 2:
Murat Basaran
Reviewer 3:
Jorge Guerra
Appl. Sci. 2022, 12(23), 12367; https://doi.org/10.3390/app122312367
Submission received: 3 November 2022 / Revised: 28 November 2022 / Accepted: 30 November 2022 / Published: 2 December 2022
(This article belongs to the Special Issue Machine-Learning-Based Feature Extraction and Selection)

Round 1

Reviewer 1 Report

Please review and improve English language throughout the manuscript.

Related works can be expanded to include more studies.

 

4.1 Data set - It would be useful to provide more information regarding the feature selection of the presented dataset.

 

4.2 Data Preprocessing - More information to be provided, regarding the creation of the vocabulary.

In several points of the documents mentioned that 'The regularization operation called dropout mechanism is added to the input layer to prevent the model from over-fitting, and attention mechanism is added to the middle layer to strengthen the feature learning of URLs.'. 

It would be useful to be added more details regarding the avoidance of the model's over-fitting.

In general more information, more details needed in Chapter 4, regarding the preprocessing and evaluation procedures. Also, it would be nice to report any limitations or drawbacks of your method as well as some lessons learned (This can be added in Section 4 or alternatively in Section 5). 

 

5. Conclusions - Future work or directions regarding the current study would be useful to be mentioned. Additionally, related domains or fields where this particular architecture could be applied, it would be better to be added in the study.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The current manuscript can be accepted as it is.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The work is interesting but there are some inconsistencies to be corrected.

The proposal shows BiGRU as the technique to be used in the proposal, however it is not explained which is the criterion to choose this model and not others.

It only talks about other techniques when comparing values at the end of the data loading, which does not explain the choice made.

only 8 of the 17 references are within the last 5 years, they should update the references and increase the number of bibliographic citations.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments have been addressed. Please consider proofreading the document to improve the use of English. 

Back to TopTop