Spider Taylor-ChOA: Optimized Deep Learning Based Sentiment Classification for Review Rating Prediction

Round 1

Reviewer 1 Report

The article may be published subject to comments (Notes.doc).

Comments for author File: Comments.doc

Author Response

Dear Sir/Madam,

We would like to express our thanks to the editor and the anonymous reviewers for their time, accurate review of our manuscript, and invaluable comments on this paper. We have carefully revised the manuscript according to the reviewers’ suggestion. All the suggestions are addressed in this response, and the corresponding changes have been incorporated in the revised paper. We hope that our revised paper can meet the requirement of publication. In the revised paper, we use the red color to indicate the changes.

1. Page 5, line 202.

The sentence “Here, the SentiWordNet feature is expressed as.” is incomplete.

Response: The required correction has been done in the revised paper.

2. Page 11.

Formulas go off the page.

Response: As suggested, the correction has been done in the revised paper.

Author Response File: Author Response.docx

Reviewer 2 Report

Dear Authors,

Thank you very much for submitting this paper. I read it with great interest, and I believe it is almost ready for publication.

Among its advantages are a clear and understandable division of the content, coherent and logical narration, well-described research process (i.e. the development of the model, the analysis of the obtained results). The research is based on a reliable, thoroughly thought-out concept, which can be implemented in practical business activities.

However, it is worth looking at your study and refining it in some areas.

Some linguistic and editing errors (spaces, punctuation) need to be corrected.

In the "Introduction", the main research goals should be stated (before discussing the division/content). I would also suggest moving the paragraph with contributions to the "Conclusions" and then expanding it, indicating the practical application of your model (i.e., assessing its implementation potential). Before the reader is familiar with the methodology and results, the current placement of contributions is too early and thus not very readable.

At the end of the literature review, it would be appropriate to derive the research questions and to demonstrate - this time more precisely - the objectives of your study (they are currently somewhat scattered, e.g. in lines 151 and 168).

In the methodology section, you could justify the choice of these three specific datasets (i.e. Yelp2014, etc.).

Most work is needed at the end. Currently, it is too superficial. You should indicate whether the objectives were met and the questions clarified. You should then present the most critical, in-depth conclusions, relating them to your research and the secondary literature. You could add an extended description of the contribution of your work to both science and practice. I would also suggest developing one final sentence about future research directions - the current statement is not substantive and does not add much to your paper. The "Conclusion" should also indicate the limitations of your research that are currently lacking.

My list of recommendations is not too long because the study itself is of good quality. However, it is worth spending some more time on it to be recommended for publication in a prestigious journal without any doubt.  

Sincerely.

Author Response

Dear Sir/Madam,

We would like to express our thanks to the editor and the anonymous reviewers for their time, accurate review of our manuscript, and invaluable comments on this paper. We have carefully revised the manuscript according to the reviewers’ suggestion. All the suggestions are addressed in this response, and the corresponding changes have been incorporated in the revised paper. We hope that our revised paper can meet the requirement of publication. In the revised paper, we use the red color to indicate the changes.

Dear Authors,

Thank you very much for submitting this paper. I read it with great interest, and I believe it is almost ready for publication.

Among its advantages is a clear and understandable division of the content, coherent and logical narration, well-described research process (i.e. the development of the model, the analysis of the obtained results). The research is based on a reliable, thoroughly thought-out concept, which can be implemented in practical business activities.

However, it is worth looking at your study and refining it in some areas.

1. Some linguistic and editing errors (spaces, punctuation) need to be corrected.

Response: We went through the entire manuscript to eliminate the linguistic and editing errors mistakes.

2. In the "Introduction", the main research goals should be stated (before discussing the division/content). I would also suggest moving the paragraph with contributions to the "Conclusions" and then expanding it, indicating the practical application of your model (i.e., assessing its implementation potential). Before the reader is familiar with the methodology and results, the current placement of contributions is too early and thus not very readable.

Response: Thanks for your comment. As per reviewer suggestion, the correction has been done in the revised paper.

3. At the end of the literature review, it would be appropriate to derive the research questions and to demonstrate - this time more precisely - the objectives of your study (they are currently somewhat scattered, e.g. in lines 151 and 168).

Response: As per reviewer suggestion, the required correction has been done in the revised paper.

4. In the methodology section, you could justify the choice of these three specific datasets (i.e. Yelp2014, etc.).

Response: As suggested, aforesaid correction has been done in the revised paper.

5. Most work is needed at the end. Currently, it is too superficial. You should indicate whether the objectives were met and the questions clarified. You should then present the most critical, in-depth conclusions, relating them to your research and the secondary literature. You could add an extended description of the contribution of your work to both science and practice. I would also suggest developing one final sentence about future research directions - the current statement is not substantive and does not add much to your paper. The "Conclusion" should also indicate the limitations of your research that are currently lacking.

Statement: Thanks for pointing out this issue. As suggested, the conclusion section is updated in the revised version of the paper.

6. My list of recommendations is not too long because the study itself is of good quality. However, it is worth spending some more time on it to be recommended for publication in a prestigious journal without any doubt.  

Statement: Thank you for your suggestion.

Author Response File: Author Response.docx

Reviewer 3 Report

Spider Taylor-ChOA: Optimized Deep Learning Based Sentiment Classification for Review Rating Prediction

1. Very interesting research entitled “Spider Taylor-ChOA: Optimized Deep Learning Based Sentiment Classification for Review Rating Prediction”.

2. Correct the structure of the article (see attached file). 

** Check "Microsoft Word template" from Applied Sciences-MDPI.

3. Use the width of the sheet for the references (see attached file).

4. Delete pages 28 and 29. No information.

5. On lines 76-77 there is a text that says “The classification of sentiment with deep learning utilize end-to-end pattern”.  I suggest referencing an article dealing with patterns. Review “A Byte Pattern Based Method for File Compression”, DOI: 10.1007/978-3-030-34989-9_10. 

6. Feature extraction is obtained as follows: “C₁ is SentiWordNet features, C₂ signifies Numerical words, C₃ is TF-IDF features, C₄ symbolize Hashtags features, C₅ is Emoticons, C₆ is Elongated words, C₇ is Punctuation marks, and C₈ is Number of sentences”. I suggest to show an example with obtained data.

7. I suggest that figure 2 be made larger (use the full width of the page); this in order that the characters in the "RMDL Model" are well appreciated.

8. On lines 308-310 it says “The CNN is adapted for categorizing the image. Despite, it is constructed to process images; the CNN is effectually adapted for classifying the texts”. There is a lot of difference between classifying images and classifying texts. How is the fit done in Convolutional Neural Networks (CNN)?

9. Equations 33-36 are not complete. I suggest using the full width of the page. (see attached file).

10. In the paragraph of lines 606-631, figures 7 and 8 are interspersed.  Figures should be inserted after the end of the paragraph.

11. In the paragraph of lines 606-631, figure 9 is intercalated. The figure should be inserted after the end of the paragraph.

12. I suggest reviewing the article “Deep neural network-based classification model for Sentiment Analysis”. DOI: 10.1109/BESC48373.2019.8963171.

13. Very good bibliography.  

The article has good content and very interesting.

Authors are requested to make all indicated corrections.

Comments for author File: Comments.pdf

Author Response

Dear Sir/Madam,

We would like to express our thanks to the editor and the anonymous reviewers for their time, accurate review of our manuscript, and invaluable comments on this paper. We have carefully revised the manuscript according to the reviewers’ suggestion. All the suggestions are addressed in this response, and the corresponding changes have been incorporated in the revised paper. We hope that our revised paper can meet the requirement of publication. In the revised paper, we use the red color to indicate the changes.

Spider Taylor-ChOA: Optimized Deep Learning Based Sentiment Classification for Review Rating Prediction

1. Very interesting research entitled “Spider Taylor-ChOA: Optimized Deep Learning Based Sentiment Classification for Review Rating Prediction”.

Response: Thanks for your comments.

2. Correct the structure of the article (see attached file). 

** Check "Microsoft Word template" from Applied Sciences-MDPI.

Response: Thanks for your comment. Sir actually I had used Latex version so I don’t know how to fix it. In my last research paper respectable editor has done that by themselves.

3. Use the width of the sheet for the references (see attached file).

Response: Thanks for your comment. Sir actually I had used Latex version so I don’t know how to fix it. In my last research paper respectable editor has done that by themselves.

4. Delete pages 28 and 29. No information.

Response: Thanks for your comment. Sir actually I had used Latex version so I don’t know how to fix it. In my last research paper respectable editor has done that by themselves.

5. On lines 76-77 there is a text that says “The classification of sentiment with deep learning utilize end-to-end pattern”.  I suggest referencing an article dealing with patterns. Review “A Byte Pattern Based Method for File Compression”, DOI: 10.1007/978-3-030-34989-9_10. 

Response: As suggested, aforesaid reference has been cited in the revised paper.

6. Feature extraction is obtained as follows: “C₁is SentiWordNet features, C₂signifies Numerical words, C₃ is TF-IDF features, C₄ symbolize Hashtags features, C₅ is Emoticons, C₆ is Elongated words, C₇ is Punctuation marks, and C₈ is Number of sentences”. I suggest to show an example with obtained data.

Response: As per reviewer suggestion, the required correction has been done in the revised paper.

7. I suggest that figure 2 be made larger (use the full width of the page); this in order that the characters in the "RMDL Model" are well appreciated.

Response: As suggested, the required correction has been done in the revised paper.

8. On lines 308-310 it says “The CNN is adapted for categorizing the image. Despite, it is constructed to process images; the CNN is effectually adapted for classifying the texts”. There is a lot of difference between classifying images and classifying texts. How is the fit done in Convolutional Neural Networks (CNN)?

Response: As suggested by the reviewer, we have revised aforesaid sentence in the manuscript.

9. Equations 33-36 are not complete. I suggest using the full width of the page. (see attached file).

Response: Thanks for your comment. Sir actually I had used Latex version so I don’t know how to fix it. In my last research paper respectable editor has done that by themselves.

10. In the paragraph of lines 606-631, figures 7 and 8 are interspersed.  Figures should be inserted after the end of the paragraph.

Response: Thanks for your comment. The required correction has been done in the revised paper.

11. In the paragraph of lines 606-631, figure 9 is intercalated. The figure should be inserted after the end of the paragraph.

Response: Thanks for your comment. The required correction has been done in the revised paper.

12. I suggest reviewing the article “Deep neural network-based classification model for Sentiment Analysis”. DOI: 10.1109/BESC48373.2019.8963171.

Response: As per reviewer suggestion, aforesaid paper has been cited in the revised paper.

13. Very good bibliography.  

The article has good content and very interesting.

Authors are requested to make all indicated corrections.

Response: We appreciate the positive feedback from the reviewer.

Author Response File: Author Response.docx

Round 2

Reviewer 3 Report

Thank you for correcting and improving the article.

Some cases were not fixed, Latex did not allow it. They are formatting details.