Bias and Cyberbullying Detection and Data Generation Using Transformer Artificial Intelligence Models and Top Large Language Models
![](/bundles/mdpisciprofileslink/img/unknown-user.png)
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis topic is very relevant, and I enjoyed very much reading this paper. However, the structure is too much disorganized. The flow is also bad. Authors need to better organize their ideas.
Improvement suggestions:
1. The goal of the study is clear but not the research gap.
2. The research questions should be aligned with the research gap.
3. The originality level of this study is not totally clear.
4. It is very important to provide a better background of the topic using more scientific references. Authors can extend introduction section or provide a literature review section.
5. The presentation of datasets should be provided in the methodology section.
6. Information provided in related work section should be better organized. It lacks flow.
7. It would be important to have a global figure that shows the several phases of your methodology.
8. Authors should have a Results section.
9. It is also crucial to have a Discussion section that you could analyze the contributions provided by your study.
10. Theoretical contributions and practical contributions should be addressed.
11. Future research directions should be aligned with the limitations of this study.
12. Figure 14 should have better quality.
Comments on the Quality of English Languagesee my comments.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe manuscript is skillfully composed. However, there are several issues that need to be addressed before the paper can be considered for publication.
1I would like to suggest that the authors consider adding the following at the end of the introduction section: use bullet points to highlight the main contributions of the paper.
2In the Project Datasets and Related Work, the details are important, but they should give an attractive story. The section is a bit long, and many of them are very general introductions,needs to be reorganized.
3Line281-285?incorrect format
4The text part of figure 7 and 14 needs to be enlarged, it's pretty blurry now
5The Validation Loss of MobileBERT is Nan,Can you discuss this briefly?
6Line454 need a space (Figure 10a) shows
7Discussion: The results and part of the discussion mostly emphasize the success of the proposed model without discussing any limitations or areas for future research. In addition, it does not sufficiently discuss the practical implications of the findings for the industry or future research directions (the end of the conclusion addresses this information -there is a suggestion to modify the text).
8Conclusion: This section needs to be rewritten to focus solely on addressing the study's objectives.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsFirstly, I commend the authors for their outstanding work in this article. The article addresses a contemporary discourse on “the effectiveness of leading Large Language Models (LLMs) in generating synthetically biased and cyberbullying data and evaluates the proficiency of Transformer AI models in detecting bias and cyberbullying within both authentic and synthetic contexts.” The work is well presented, and a step-by-step methodology is outlined with some good diagrammatical evidence of the procedure adopted. The title is precise, while the research questions and the aim of the study are clearly outlined.
However, there are a few areas that should be addressed to strengthen the article:
1). Line 51-52: You stated, "The rapid proliferation of synthetic data generated by advanced AI systems has intensified the need to address biases inherent in such models". Who said this? This statement should be substantiated with supporting references.
2). Page 4 of 29: Line 40-48: You need to explain clearly what happened or the implication of your finding in Figure 1. If the intention was to “generate copyrighted content” and “just a sentence from a Tina Turner’s song was generated” was this mission not partially achieved? What does this mean about the efficiency of the copyrighting mechanism used? From your explanation given, these questions are not answered.
3). Line 192: “As can be seen from the table, negative tweets…” Your article has got many tables, its best that you sequentially refer to each table by its numbering as you have been doing in Tables 1 to 4 above. Consistence in your presentation is key throughout your article.
4). The methodology is precise, and all the algorithms are well explained and supported with generated tables and graphs, thus, this section is well articulated. I have got no issues.
5). Lines 597 - 616: What are the contributions of your study to literature, managerial practice, and legal/policy within the legal fraternity and the cyberbullying space? What were the limitations of your study? Your future research direction is clearly stated. This missing information should be added.
The rest of the article is okay.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsI recognize some good improvements, but the discussion of results must be extended and based in scientific evidence. Furthermore, the organization of the Conclusions section is very poor and strange.
Improvements:
1. It is important to organize the studies presented in Table 1.
2. Discussion section needs to be significantly extended to demonstrate the main differences regarding previous studies. It is important to support your discussion and the novelty and relevance of your studies in a scientific background.
3. I do not think that it is a good idea to present the contributions and future research directions into two tables. Instead use plain text and create two sections for that.
4. Articulate your future research directions with the current limitations of your work.
5. I also recommend increasing the number of references. This topic has attracted a lot of scientific interest in recent years/months.
Comments on the Quality of English Languageit is ok. but some formatting issues appear.
Author Response
Please find our answers in the file attached.
Author Response File: Author Response.pdf
Round 3
Reviewer 1 Report
Comments and Suggestions for AuthorsI recommend the authors to better structure the Conclusions section. Please consider my suggestion:
6.1 Final remarks
6.2 Theoretical contributions
6.3 Practical contributions
6.4 Future work
Author Response
Thank you for your suggestions below:
I recommend the authors to better structure the Conclusions section. Please consider my suggestion:
6.1 Final remarks
6.2 Theoretical contributions
6.3 Practical contributions
6.4 Future work
Section 6. Conclusions and Future Work was restructured as suggested.