A PTM-Based Framework for Enhanced User Requirement Classification in Product Design

Round 1

Reviewer 1 Report

This is a structured paper of interest in the context of the publication. The objectives of the study are clear as well as the introduction and declaration of problems.

The Introduction, problem statement and objective of this paper is clear. Acronyms and abbreviations should be reviewed, e.g. natural language processing (NLP), needs to be Natural Language Processing (NLP).

The Related Works is supported by pertinent and recent bibliographical references. Almost all the references are from the last five years. The related work could have been presented in more detail given the importance of the issue under study.

The Method described and detailed enough to understand the approach. Figures 3  - Fine-tuning SKEP model structure, should have a higher level of legibility.

The Experimental Settings section presents the data collection assumptions and methods underlying the study in sufficient detail to understand the approach.

The Experimental Results and Analysis section report and present the experimental results of the Requirements. The Figure 4. Visualization of clustering Effect of diferente Requirement topic models, should have a higher level of legibility.

The Conclusion should relate the method, experimental settings, experimental results and analysis. Authors should clearly present the strengths of their work. It should clarify the added value of the paper as well as the limitations and the prospect for future research.

Improving the quality of the English language

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

To the authors of the publication, I have a few tips that I hope will improve the meaningfulness of this article.

1.    The authors did not identify the drawback of existing new methods for requirement classification. The motivation of the proposed approach is not clear.

2.  The cited references are relevant to the research and there is almost no self-citation. The authors improve the literature review of the process of PTM-based framework for enhanced user requirement classification.

3.    The Introduction does not show the research gap. Please highlight the main objectives of the proposed of PTM-based framework for enhanced user requirement classification.

4.  On Figure 4 the authors visualize the clustering effect of each model, except LDA. Give a reason for not visualizing this model.

5.  In section 4.1 you stated the number of user's requirement attributes K = 14. Nowhere is this number explained. Tell us which ones they are and why you chose them?

6.  In Table 1, explain the Sil Score values. Nowhere is it stated whether higher or lower Sil Scire is preferable.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf