A Secure Live Signature Verification with Aho-Corasick Histogram Algorithm for Mobile Smart Pad
Round 1
Reviewer 1 Report
The article is interesting.
It is well written.
Unfortunately, it has several disadvantages:
1. There is no scientific experiment design.
2. There is no statistical analysis of the results.
3. It is not known what is the statistical significance of the obtained results.
The presented conclusions, in my opinion, are not confirmed by the material presented in the article.
E.g:
L. 607 "...Most of the accuracy is below 50%. The total accuracy is only 46%..." This sentence is a mystery to me. What does it mean?
L. 617 "...Most of the accuracy is above 50%. ..." This is not a scientific statistical analysis. 50% accuracy? What does that actually mean?
On item 6. Conclusions there are no conclusions ... there are general remarks.
I have doubts whether the article is a scientific text.
Author Response
Dear Reviewer:
Please see the attachment. Thanks.
Best Regards
Charles
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This paper proposes a handwriting signature verification algorithm by using four live waveform elements as the verification features. A new Aho Corasick Histogram mechanism is proposed to perform this live signature verification.
The paper is well organized and readable. I have some suggestions, which are described below, to be considered to improve the paper.
First, the references are old. The most recent references are from 2019. At least some references from 2020 are missing. It follows that related work (Section 2) has died out after 2019.
The results of the experiments are extensive, but in my opinion, the results are misrepresented. With the additional experiments should be confirmed that the proposed method is excellent. It is necessary to compare it with other methods. Use the methods you presented in the second section (Related work). It would be best to do comparative experiments with the latest algorithms in this field. It would be a good idea to also use some established signature database.
I recommend that the paper should be accepted with major revision.
Author Response
Dear Reviewer：
Please see the attachment. Thanks.
Best Regards
Charles
Author Response File: Author Response.pdf
Reviewer 3 Report
The authors have proposed a secure live signature verification method. Please see the comments to improve the paper further.
- Kindly proofread the paper to omit many typos and grammatical mistakes.
- In the abstract, it is not clear why do new scheme is needed? A problem statement is missing. Further, mention quantitative improvements to show the advantages of the proposed scheme compared to the other.
- The paper falls under the area of multimedia IoT and it should be introduced with this recent reference. https://doi.org/10.1109/ACCESS.2020.2964280
- The limitations of existing schemes should be mentioned before introducing the proposed method to show how it caters to these limitations in the introduction Section. Extend the Introduction section to introduce the research area well.
- The second paragraph of the related work seems to be the paper's contribution and should be part of the Introduction section. Rewrite the whole related work and include the relevant papers only.
- Include table of notations used in the equations to increase the readability of the equations.
- Add physical meaning of the equations.
- Generically write pseudo-code and then explain it line by line in the body text.
- Add detail experimental parameters in a tabular form.
- The proposed scheme should be compared with the most recent method to show the effectiveness. It is compared with only the standard methods.
Author Response
Dear Reviewer:
Please see the attachment. Thanks.
Best Regards
Charles
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
This paper proposes a handwriting signature verification algorithm by using four live waveform elements as the verification features. A new Aho Corasick Histogram mechanism is proposed to perform this live signature verification.
The paper is well organized and readable. I have some suggestions, which are described below, to be considered to improve the paper.
The authors partially complied with my recommendations in improving the paper. References have been supplemented.
My opinion is that the papers in this journal with a good impact factor should also present comparative experiments performed with other algorithms in the field. These comparison experiments should be done with the best algorithms in the field of the algorithm presented.
Since the authors did not do comparative experiments, unfortunately, I cannot suggest the paper to be accepted in this journal.
Author Response
Dear Reviewer:
Thanks for your comments. We strongly agree with reviewer’s comments. The main contribution of this research is to propose a brand-new ACH algorithm with good usable accuracy. In our experiment, we have a fair comparison with Euclidean Distance (see section 5.1, line 641~651)、Correlation Classification Algorithm (see section 5.2, line 652~661)、Convolutional Neural Network (CNN) (see section 5.3, line 662~668) and Deep Learning Algorithm (see section 5.4, line 669~681), which should prove the practicality of our algorithm. We also explain why the average accuracy rate of CNN and VGG16 are worse than proposed ACH algorithm.
Best Regards
Charles
Author Response File: Author Response.pdf
Reviewer 3 Report
I am satisfied with the revised manuscript.
Author Response
Dear Reviewer: We very appreciate your instruction and comments in revising paper. Thanks again.
Best Regards
Charles
Round 3
Reviewer 2 Report
This paper proposes a handwriting signature verification algorithm by using four live waveform elements as the verification features. Furthermore, a new Aho Corasick Histogram mechanism is proposed to perform this live signature verification.
The paper is well organized and readable. Moreover, the authors complied with my recommendations in improving the paper. Therefore, I have no more suggestions.
I recommend that the paper should be accepted.
