Systemic Acquired Critique of Credit Card Deception Exposure through Machine Learning

Round 1

Reviewer 1 Report

·         The English needs improving.

·         The formatting needs improving, often words are adjacent, with the space missing, including the title.

·         Figure 4 should be a bar chart.

 

·         The paper does the groundwork well, which is clearly explained.  What is missing is drawing significant and interesting conclusions.  How effective is the application of machine learning to the detection of credit card fraud?  Which methods work best?  Any prescriptions or recommendations for the future?

Author Response

The English needs improving. The formatting needs improving, often words are adjacent, with the space missing, including the title.	Agreed, Done. We have improved English. Formatting is improved the space is adjusted. Title is adjusted too Line 1- 399
Figure 4 should be a bar chart.	Agreed, Done. Line 295
What is missing is drawing significant and interesting conclusions.  How effective is the application of machine learning to the detection of credit card fraud?  Which methods work best?  Any prescriptions or recommendations for the future?	Agreed, Done. Conclusion is updated. Line 366-386

Author Response File: Author Response.docx

Reviewer 2 Report

Fraud detection by applying novel ML techniques is quite popular & well researched in the community. Numerous survey papers exist already. In fact, there are ML driven solutions to perform such surveys in an automated way. So, I am not sold on the motivation of the authors to come-up with a seemingly manual system to survey such works now. Also, it is not entirely clear to me if the system proposed and followed by the authors here is done manually or programmatically. Such details seem to be missing. Finally, there are quite a few English / Grammar related issues throughout the article (hard to list them all). 

Having said the above, the results provided here seem quite extensive and worth sharing to a broader audience. My recommendation to the authors is to get the article proof read by someone who is well versed with scientific publications & resubmit.

Author Response

Fraud detection by applying novel ML techniques is quite popular & well researched in the community. Numerous survey papers exist already. In fact, there are ML driven solutions to perform such surveys in an automated way. So, I am not sold on the motivation of the authors to come-up with a seemingly manual system to survey such works now.	Agreed, Done. We provided . line 146-154 We provide a clear research gap. What is the difference between existing research and this research.
Having said the above, the results provided here seem quite extensive and worth sharing to a broader audience. My recommendation to the authors is to get the article proof read by someone who is well versed with scientific publications & resubmit	Agreed, Done. Proof reading is done by a researcher who  well versed with scientific publications.

Author Response File: Author Response.docx

Reviewer 3 Report

Introduction should be more elaborate

Related works can be added as a separate section.

Quality of figures is so important too. Please provide some high-resolution figures. Some figures have a poor resolution.

What is the motivation of the proposed work? Research gaps, objectives of the proposed work should be clearly justified.

 

 

 

Author Response

Introduction should be more elaborate	Agreed, Done. Introduction is elaborated according to the suggestion. line 38-48 Line 88-98
Related works can be added as a separate section.	Agreed, Done. Related work is added as a separate section. Line 102-142
Quality of figures is so important too. Please provide some high-resolution figures. Some figures have a poor resolution.	Agreed, Done. Quality of figures is improved.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

The overall read is much better now, thanks for rewriting a majority of the article. Tables 2,3,4 are quite hard to read and comprehend due to the extensive amount of information in there. Is there a way to better represent that data? Your survey is comprehensive but the summary of your work can be improved. For example, is there a clear winner or top few combinations of  techniques, datasets, performance metrics that can be suggested as a result of this extensive survey?

Author Response

Thank you for re-submission and for addressing my previous concerns.	Thank you!!
Questions:
Table 2 i quite hard to read and comprehend due to the extensive amount of information in there. Is there a way to better represent that data?	Agreed, Done. Table 2 is updated according to the suggestions. Line 303-310
Table  3 is quite hard to read and comprehend due to the extensive amount of information in there. Is there a way to better represent that data?	Agreed, Done. Table 3 is updated. Line 321-328
Table 4 is quite hard to read and comprehend due to the extensive amount of information in there. Is there a way to better represent that data?	Agreed, Done. Line 334-342
Your survey is comprehensive but the summary of your work can be improved	Agreed, Done. The summary is updated Line 382-389 Line 395-401

Author Response File: Author Response.docx