Artificial Intelligence Classification Model for Modern Chinese Poetry in Education

Round 1

Reviewer 1 Report

1. It had better for the authors to ask a professional English editor to revise the manuscript.

2. In lines 67 and 68, the authors mentioned the XGBoost-MCP model is obviously superior to the other three algorithms. The authors should clearly explain what algorithms are the other three algorithms.

3. The literature citation location looks weird; the literature citation location should be after a paragraph that related to the cited literature, instead of after the authors of the cited literature.

4. Some abbreviations should have their full texts when they appear first in the manuscript.

5. The authors should have more descriptions to explain the labeling step in the data preprocessing of their proposed prediction framework; especially how to label a style of a poem.

6. It had better for the authors to have a legend to explain the symbols in Figure 2.

7. The authors should explain the meanings of the symbols in equation 1, equation 3, equation 4, and equation 5.

8.  It had better for the authors to have a symbol table to explain the meanings of all symbols in Subsection 3.3.

9. The authors had an introduction to the XGBoost algorithm in Subsection 3.3. Why did not the authors have introductions of the other algorithms?

10. The authors should explain the reasons why they used the four models, XGBoost-MCP, SVM, DNN, and DT, in the experiment to evaluate the performance of those classifiers.

11. Why did the authors use two schools in a test set, instead of three schools or four schools?

12. Regarding the experiment result analysis, besides finding the best performance among the four models, the authors also can explore the performance of the other models in a modern Chinese poetry style classification.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

What does this paper contribute to sustainability? This paper seems more suitable for computer science (Journal of Theoretical and Applied Electronic Commerce Research) or AI-affiliated journals (AI). To be published in Sustainability, a paper's research, scope, and findings must be suitable for the journal. This study does not mention, claim, or connect related to sustainability in any part of the manuscript.

Other comments

It is difficult to see the following contents as contributions to this paper, so it would be better to replace it with other actual contributions.

After labeling, word segmentation, and removal of stopwords, the text data of modern Chinese poetry are preprocessed, Doc2Vec and XGBoost algorithm is used to iteratively train the XGBoost-MCP model.

What is the basis for the data classification ratio (753:83)?

Who would benefit from a better classification of poetry? Academically, it is acknowledged that the calculation accuracy has increased. However, what concrete help does this research have for mankind in terms of learning and education?

The manuscript should be meticulously proofread by a native speaker.

Wordy: The experimental results show that the XGBoost-MCP model is obviously superior to the other three algorithms and has high accuracy and objectivity, and applying it to education can help learners and researchers better understand and study poetry.

Grammarly wrong: The Modern Chinese Poetry based on XGBoost (XGBoost-MCP) model is built in this paper makes poetry classification more accurate, objective and efficient, it solves the defects of traditional poetry education.

Grammarly wrong: After labeling, word segmentation, and removal of stopwords, the text data of modern Chinese poetry are preprocessed, Doc2Vec and XGBoost algorithm is used to iteratively train the XGBoost-MCP model.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

It is very interesting research on how understanding poetry styles can be done using AI. However, the paper still needs some revision for consideration. 

1. It is good to provide quantitative evaluation of findings for the conclusion in abstract. However, overall, the abstract is very good. 

2. Sentence in Line 60-69 is recommended to be in conclusion.

3. Theare so many repetitive sentences that begin with "In xxxx (year)...". Kindly rewrite so that the writing will not bore the readers. Kindly check through the whole manuscript.

4. Under heading - Experiment. Dinzhi Lan was mentioned as refenced data. There was no reference or discussion on this in the introduction.  It is an important part of the work for the validation. How is the validation being done in the data training?

5. How equations 15-18 being derived? Source?

6. Any mistake in the content of table 3 and 4 (the same data)

7. Line 470 - 473. need to have some explanations why XG Boost better. Factors? Mechanism? Results and discussion also require validation from previous study. It can be done by just providing citation from papers you cited in the intro.

8. The references are good covering related papers and with an appropriate current year's coverage. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

All comments are revised and answered. 

It had better arrange Table 1 at the beginning of Subsection 3.3.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors sincerely responded to my comments and made significant revisions to the content. However, they only answered in the cover letter why this study was suitable for sustainability, but did not describe it in the paper. In the process of revising the paper, it is not enough for them to respond only through the answer form. In particular, the authors did not even reflect one of the papers cited in the response in the revised version. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf