Study on Dynamic Evaluation of Sci-tech Journals Based on Time Series Model

Round 1

Reviewer 1 Report

The paper titled “Study on Dynamic Evaluation of Sci-tech Journals Based on 2 Time Series Model” discusses dynamic evaluation of sci-tech journals.

The authors have not listed the 18 journal evaluation metrics mentioned in the abstract. The abstract should be structured, it should highlight the methodology (at least in brief) and result analysis (statistics or numbers or improvement in accuracy).

Authors have cited recent publications (2021) in their paper.

The authors have discussed their proposed work exhaustively which makes this paper a good read.

The authors have demonstrated the high prediction accuracy of their proposed method. They have proposed dynamic evaluation tasks for evaluating journals based on proposed evaluation metrics. The proposed model can be extended for future evaluation of journals. The proposed methodology with the help of datasets generated can be used for similar studies.

Tables and Figures presented in the paper are appropriate and convey the message without reading the entire manuscript. The proposed methodology is sound, and results corroborate the arguments proposed by the authors. The paper has archival value and will help future research in similar areas.

 No specific comments regarding the manuscript or methods.

Reviewer 2 Report

The authors of the submitted manuscript first justify their motivation to investigate the issue in the abstract. They also describe the content and methodology of their contribution here, and they outline other possibilities for investigating the issue. I can recommend to the authors of this manuscript to briefly present the findings and results in the text of the abstract.

The paper has a logical structure and organization, the individual parts follow each other smoothly. The authors' interpretation is fluent, clear, and understandable. The manuscript contains enough references to sources from professional literature. All figures in the text are of high quality and have great explanatory value. Tables should not extend beyond the page edges. All mathematical formulas are written very well. Formulas, tables, and figures are properly numbered. The content of the contribution is beneficial from the point of view of scientific knowledge.

The introductory chapter of the submitted manuscript represents the authors' own motivation to deal with the given topic. The authors highlight the importance of the given area of research, emphasize the need for research in this area and state the limitations of research on the given question. The introductory chapter also includes a high-quality analysis of the current state of knowledge. The introductory chapter further presents the content of the submitted manuscript and presents the methods and procedures used by the authors in their research. The authors also present here the results of their research and the possibilities of their use. I can recommend to the authors of the manuscript also to pay attention to the specification of the research objectives in the introduction of the paper and further the formulation of the research hypotheses that they will try to prove during their research. In the introductory chapter, the authors present the content of their contribution and its structure.

In the conclusion chapter, the authors of the manuscript emphasize their own contribution to the addressed issue and present the results and findings they have reached as part of their research. They indicate the possibilities of expanding their research and practical application of the already obtained results. In this chapter, the authors of the submitted manuscript provide a final summary of the knowledge obtained.

Reviewer 3 Report

The paper presents a proposal for a dynamic evaluation of Sci-Tech Journals based on time series models trained by machine learning techniques.

The paper is well written and structured.

However, in the following, several concerns/questions that the Authors have to give answers in text (i.e., to better clarify the reader doubts) are pointed out by this Reviewer:

1. In formula (1), there is a bias by summing the "minimum values of the mapping interval", Min (see the last term). Why this is necessary? Give a written explanation in text.

2. The ANN in Figure 5, does the output layer has bias vector of coefficients? 

Best Regards,

The Reviewer

 

Reviewer 4 Report

1. Indicate the findings in the abstract.

2. Inform about your novelties in the introduction. Also, highlight your research questions.

3. Subsections 2.2.1 and 2.2.2 are not related to this paper.

4. Explain why you didn't prefer well-known normalization techniques like vector, max-min, etc. Why did you use the method in Eq. 1? Who did propose it?

5. A language and writing check is needed.

6. Cite also all performance measurements such as MSE, MAE, ...

7. Discussion is poor. It is expected to compare your results with other papers.

8. Theoretical and practical implications must be added as a sub-section.

9. Limitations haven't been discussed.

10. Conclusion part is long. Briefly discuss the significance of the study and its findings. Contributions, future studies, and limitations should be highlighted.

11. The literature review is quite insufficient for such a comprehensive study. You should discuss studies that use methodologies such as ANN and deep learning. You are expected to summarize the methods (MLP, RF, CNN, ...) in different fields of science with a table. The following recent papers are also recommended.

How can we manage offensive text in social media-a text classification approach using LSTM-BOOST. 

Machine learning models for predicting the residual value of heavy construction equipment: An evaluation of modified decision tree, LightGBM, and XGBoost regression. 

Training multilayer perceptron with genetic algorithms and particle swarm optimization for modeling stock price index prediction. 

LSTM integrated with Boruta-random forest optimiser for soil moisture estimation under RCP4. 5 and RCP8. 5 global warming scenarios.