Application of Machine Learning and Multi-Dimensional Perception in Urban Spatial Quality Evaluation: A Case Study of Shanghai Underground Pedestrian Street

Round 1

Reviewer 1 Report (Previous Reviewer 1)

Comments and Suggestions for Authors

Comments and Suggestions for Authors:

1. Methodological Transparency (Lines 131-198):

• Comment: The updated methodology section improves understanding but still lacks some specifics on how different data types are integrated before analysis.
• Suggestion: Include a detailed flowchart or step-by-step breakdown that visually represents the integration of machine learning with multidimensional sensory data. This would not only enhance clarity but also aid in the reproducibility of your research.

2. Practical Implications (Lines 479-496):

• Comment: The conclusion effectively underscores the study's contributions but could be enhanced by discussing specific urban planning applications in more detail.
• Suggestion: Expand on how municipalities or urban planners might implement your findings in practical scenarios. Discuss potential challenges and solutions in applying these methods to different urban environments, which would provide readers with a clearer pathway from theory to practice.

3. Technical Jargon and Accessibility (Lines 200-205):

• Comment: You have made strides in clarifying technical terms; however, ensuring that the manuscript remains accessible to a broad audience, including those outside your immediate field, remains crucial.
• Suggestion: Continue to define technical terms at their first occurrence and consider adding a glossary of terms if the manuscript continues to use specialized vocabulary extensively.

4. Further Proofreading (Lines 230-235):

• Comment: Although improvements have been noted in the language quality, ensuring the manuscript is free from typographical and grammatical errors is essential for maintaining professional standards.
• Suggestion: It may be beneficial to engage a professional proofreading service to conduct a final review of the manuscript to ensure language precision and coherence.

5. Future Research Directions:

• Comment: The manuscript hints at future research directions, but these could be more explicitly linked to the current findings.
• Suggestion: Detail specific future research questions that arise directly from your findings. Discuss how your methodology could be adapted for above-ground urban spaces or integrated with emerging technologies like AI for predictive modeling. This approach will not only enrich the discussion but also open up avenues for further scholarly exploration.

Comments on the Quality of English Language

I explained it in the suggestions for the authors.

Author Response

To address Reviewer 1's concerns about the clarity of data integration, we included a detailed flowchart and a step-by-step explanation of how machine learning is integrated with multidimensional sensory data. This addition enhances the transparency and reproducibility of our research. Furthermore, we expanded the discussion to include practical implications, detailing how urban planners can apply our findings in various urban environments and outlining potential challenges and solutions.

Thank you sincerely for your positive feedback and recognition of our work. Your encouragement is a strong driving force for us to continue our efforts. We are committed to contributing to the development of sustainable and high-quality urban spaces and aim to produce more impactful research in the future. We look forward to further engaging in meaningful discussions with you.

Please read the attached for details:

Author Response File: Author Response.docx

Reviewer 2 Report (Previous Reviewer 2)

Comments and Suggestions for Authors

Dear authors, 

I am honored to review this research entitled  "Application of Machine Learning and Multidimensional Perception in Evaluating Urban Space Quality: A Case Study of the Underground Pedestrian Street in Shanghai." The research topic "Application of Machine Learning and Multidimensional Perception in Evaluating Urban Space Quality" and "Wasted Spaces and Traffic Congestion in Today's Underground Pedestrian Streets" present promising research areas. Although the evaluation indicators are not new, the proposed model integrates spatial measurement techniques, machine learning algorithms, and multidimensional perception metrics, offering novelty in this research field. However, the research needs some methodological and thematic issues. Therefore, several points can be considered to improve this research:

1. This research can be classified under the "Research Methods" scope, which it systematically lacks. All used methods should be presented in all research sections, or the research should build its idea on developing a measurement methods model. Therefore, the authors should briefly mention these methods in the abstract and introduction, linking them to the framework that should align with the literature review. Then, in the research design (or as they call it, methodology), they should mention the relationship between the methods and the study steps to develop the framework presented in Figure 1 from the literature review.
2. The literature review lacked focus on the primary research context, "Today's Underground Pedestrian Streets," and the implications of the methods used based on evidence from recent sources.
3. The examples discussing research areas "Machine Learning" and "Multidimensional Perception" should have been directed towards today's underground pedestrian streets, making the study gap more apparent.
4. The literature review needs an introductory paragraph explaining why the focus is on the three mentioned points in the review and not others. The literature review section should focus on the research context and not delve into side points that may have existed under different circumstances. For instance, the first paragraph discusses modernism and parametric measurement and their relation to the research context of "Today's Underground Pedestrian Streets." The historical perspective needs a solid argument to link it to the current study topic, which is the role of the introductory and concluding paragraphs in the literature review.
5. There is a methodological and logical disconnect between the literature review and the rest of the research. The literature review section should have concluded by hinting at a prototype "integrating spatial measurement techniques, machine learning algorithms, and multidimensional perception metrics" from the literature review, highlighting the deficiencies and drawbacks, followed by the researchers presenting their proposal and contribution to this model.
6. This proposed framework (Figure 1) should have emerged from the literature review and should not be placed at the beginning of the methods section.
7. The methods section should start with the case study and then proceed with the research design, which includes the steps the researcher will follow.
8. There is an issue with the numbering of (Figure 1) on page 5, line 204; the authors mean Figure 3. This mistake requires rearranging the figures with the context.
9. The methods section needs to be rewritten to show the methods used in the research, and then the research design needs to illustrate the steps of using these methods to reach a logical conclusion. Note that some methods appear for the first time, such as OSM, on page 5, line 208.
10. The research method used should be linked to the current research context. For example, what is the purpose of using The DBSCAN on page 6, line 225? Therefore, the research design must be structured in stages and steps with specific objectives and functions rather than generally and vaguely described. Also, why were these methods, such as the literature analysis method and the Delphi method, not mentioned in the literature review to build the general framework of the research?
11. I need help finding the proposed model promised by the researchers. Building the model requires different methodological thinking. Starting with a detailed framework that includes gaps, experimental and field studies, and exploring opinions can lead to a model that other researchers can apply.

Author Response

Reviewer 2 noted confusion in the discussion section, which led to scepticism regarding the study’s conclusions. We have revised this section to provide a more precise comparison with existing literature, explicitly explaining the differences in findings arising from our unique methodology and the incorporation of new evaluation indicators. The conclusion has been restructured to emphasise the study’s contributions beyond conventional knowledge, focusing on the innovative aspects of our approach.

We sincerely appreciate your thoughtful suggestions and unwavering support. Your detailed feedback has been invaluable, and We have carefully considered and implemented all of your recommendations to enhance our work. This experience has been incredibly enriching for us, and We are grateful for the opportunity to learn and grow from your insights. It is an honor to have received your guidance, and We extend our heartfelt thanks for your time and effort.

Please see the attachment for detailed changes！

Author Response File: Author Response.docx

Reviewer 3 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

Although the thesis has undergone some revisions, there are still many problems.

(1) Inadequate writing of the abstract, confusing logic and doubtful innovation

Firstly, the research methodology and conclusions are not introduced in the abstract, and it is unclear what the paper's actual innovation and research significance are. The authors claim that there are very few spatial evaluation studies using human perception data. Yet, there have been so many such studies in the past two years that even the collection and analysis of socio-perceptual data has become a prevalent means of analysis.

Further, in the abstract, the authors mention that most of the current research analyses spatial data through machine learning, indicating that machine learning is not the innovation of this study, yet the application of machine learning in spatial evaluation is emphasized in the title, so why emphasize it when this research is very common?

In the penultimate sentence, this study integrates spatial quantification methods, machine learning algorithms, and multidimensional perceptual measurements. Why does the title again not reflect spatial quantification methods, but only machine learning and multidimensional perception?

It gives a very confusing feeling and the logic is not clear.

(2) Scattered research methods

The goal of the thesis is to evaluate the perception of the underground pedestrian street, but not all of the research methods are centered on this goal. The article uses spatial syntax to analyze, which is not related to the perception of the pedestrian street, does not directly produce direct research results, and at most counts as a part of the research preparation.

(3) Discussion

The description of the conclusions in the Discussion section is very confusing and contradictory, and the reviewers could not clearly understand the differences and contributions between the main conclusions of this paper and the existing literature.

It does not explain the reasons for the differences between these conclusions. Is it because of the particular research methodology used? Were new evaluation indicators incorporated?

(4) Conclusion

The confusion in the discussion section leads to scepticism about the concluding findings, which are currently not beyond conventional knowledge.

Overall, it feels that this revision was very rushed and needs to be done in a sedate and serious manner, rather than in a reactive manner.

Moreover, this paper is not suitable for viewing in the journal Land, and I believe that there is a discrepancy with the scopes of this journal.

Author Response

In response to Reviewer 3's feedback, we clarified the research methodology and highlighted the innovative aspects of our study in the abstract. We restructured the introduction to present the research goals, emphasising integrating spatial quantification methods with machine learning and multidimensional perception. This restructuring aims to establish a logical flow that connects the literature review, methodology, and results.

Once again, we would like to thank you for your in-depth analysis and valuable feedback during the review process. Your high standards and meticulous review have helped us identify areas for improvement in our paper, and we are deeply grateful for this. With your suggestions and feedback, we believe our research will become more refined and impactful. We sincerely appreciate your support of our work and extend our highest respect to you. Your critiques and suggestions are viewed as vital forces driving academic progress.

Please see the attachment for detailed changes.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report (Previous Reviewer 2)

Comments and Suggestions for Authors

After carefully reviewing the manuscript, I noticed it has undergone significant improvements and is now ready for publication with minor revisions. I recommend enhancing the figures and graphics to ensure clarity and visual impact. Additionally, incorporating references from other countries could enrich the manuscript and contribute to a more diverse scientific discourse.

Reviewer 3 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

Congratulations, the authors have thoroughly revised the article 

My requirements have been met.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

General Overview:

Your manuscript introduces an innovative approach to evaluating urban underground pedestrian streets through a multidimensional perception-based framework. Integrating spatial syntax analysis, sensory perception measurements, and machine learning offers a comprehensive methodology that is novel and highly relevant to current urban planning challenges.

Detailed Comments

1. Introduction and Contextualization (Lines 28-43):

   • Comment: The introduction effectively sets the stage for the study by highlighting the need for better-designed urban underground spaces. However, it could benefit from a more direct linkage to the broader context of urban transitions and regeneration.
   • Suggestion: Consider incorporating references that discuss the role of data-driven methodologies in urban regeneration to strengthen your theoretical framework. For instance, "Networking analysis in the urban context. Novel instrument for managing the urban transition" could provide valuable insights into how network analysis and data-driven approaches are increasingly critical in managing complex urban transitions. This reference could help situate your study within the larger discourse on sustainable urban development.

2. Methodology (Lines 131-198):

   • Comment: Your description of the methodology is detailed and methodologically sound. However, integrating various data types and the transition between analytical methods could be more precise.
   • Suggestion: Elaborate on how the data from different sensory perceptions are integrated and analyzed within your machine learning models. A more detailed workflow or a diagram could significantly enhance the reader's understanding of the methodological transition from raw data to actionable insights.

3. Results (Lines 264-342):

   • Comment: The results are intriguing and well-presented, but there is room for more profound analysis regarding the implications of these findings on actual urban planning practices.
   • Suggestion: Expand on how these findings can be applied in the planning and redesign of other urban spaces. Specific examples or case studies where sensory data has informed successful urban regeneration projects would be very illustrative and beneficial.

4. Conclusion (Lines 479-496):

   • Comment: The conclusion effectively summarizes the research findings but could further emphasize the practical implications and potential for scalability.
   • Suggestion: Discuss how the methodologies and findings of this study could be adapted or scaled to other urban contexts or types of spaces. Highlighting your approach's versatility could broaden your work's impact and provide a more precise roadmap for its application in diverse urban settings.

Additional Recommendations

• Reference Utilization: Incorporating the suggested reference on networking analysis could enhance the theoretical depth of your paper and position it within a cutting-edge field of urban study. This could increase the relevance and citation potential of your work.
• Future Research Directions: Suggest potential areas for future research that build on your findings, such as applying the methodology to above-ground urban spaces or integrating more advanced AI techniques for predictive modeling.

Final Thoughts: Your manuscript significantly contributes to urban planning literature, especially in the under-explored area of underground pedestrian environments. With the proposed enhancements, it has the potential to influence a wide range of future urban projects, contributing to more livable and human-centered urban environments.

Comments on the Quality of English Language

The manuscript is well-written and has a clear academic tone for a scholarly audience. However, to enhance the clarity and professionalism of the text, there are several areas where the English language usage could be improved:

1. Sentence Structure and Clarity (Lines 88-94):

   • Issue: Some sentences are overly complex, making them difficult to follow. This complexity can obscure critical points.
   • Suggestion: Simplify these sentences to improve readability. Break long sentences into shorter, more concise ones. For example, instead of "The methodology that was applied was beneficial in that it allowed for a comprehensive understanding of the different elements," consider "The methodology provided a comprehensive understanding of different elements."

2. Consistency in Terminology (Lines 107-112):

   • Issue: Inconsistencies in using specific technical terms can confuse readers. For instance, "urban underground spaces" and "subterranean pedestrian pathways" are used interchangeably without clarification.
   • Suggestion: Choose one term and stick with it throughout the manuscript, or clearly define each term when it is first used to ensure that readers understand they are synonymous.

3. Grammar and Punctuation (Lines 150-156):

   • Issue: There are minor grammatical errors, such as issues with subject-verb agreement and the use of incorrect prepositions.
   • Suggestion: Review the manuscript for grammatical accuracy. For example, "The data was collected" should be "The data were collected," as " data " is typically a plural noun in academic writing.

4. Technical Jargon and Accessibility (Lines 200-205):

   • Issue: The manuscript occasionally uses jargon that may not be accessible to all readers, especially those new to the field.
   • Suggestion: Provide definitions for highly technical terms or replace them with more commonly understood language when possible. This will make your paper more accessible to a broader audience, including those outside your immediate field.

5. Proofreading for Typographical Errors (Lines 230-235):

   • Issue: There are occasional typographical errors and misspellings throughout the text.
   • Suggestion: Conduct a thorough proofread of the manuscript or employ a professional language editing service to correct these errors. This will enhance the overall quality and readability of the paper.

Addressing these points will improve the language quality and ensure that the manuscript communicates its valuable insights more effectively. This is crucial for broader dissemination and impact within the academic community.

Author Response

We have made comprehensive changes to the article based on your suggestions and answered all the queries you raised, as detailed in the document:

Reviewer 2 Report

Comments and Suggestions for Authors

Dear authors,

Thank you for allowing me to evaluate your research paper, “Evaluation of the Spatial Quality of Urban Underground Pedestrian Streets Based on Multidimensional Perception.” Despite the effort put into this paper, it currently reads as a collection of disjointed topics lacking cohesive scientific logic. Many aspects need careful revision before being considered suitable for review, let alone publication.

First, the title needs to be more specific and fully expressive. The current title, “Evaluation of the Spatial Quality of Urban Underground Pedestrian Streets Based on Multidimensional Cognition,” is too general and does not clearly define the scope of the study. It fails to specify the context in which the evaluation will occur or the research method used. While the title should be brief, it must also be precise and engaging to attract readers.

Second, the abstract needs to be rewritten professionally and scientifically, avoiding terms like “Background” and “Methods.” The language should be more robust, coherent, and fluent. Additionally, the abstract should present the problem, its context, and the research gap that underscores the study’s significance.

Third, there is a technical error: none of the references appear in the text, making it impossible to review them reliably.

Fourth, the introduction needs to be rewritten to follow a scientific structure: general introduction, study problem, gap, solution to the problem, objective, and contribution. The current introduction reads more like a literature review of the researcher’s method. It is preferable to briefly introduce the problem and its solution in the introduction and then transfer the detailed literature review to a separate section on the theoretical background. This section should culminate in a theoretical framework applicable to the case study.

Fifth, if the researcher claims that Spatial Syntax is the appropriate approach to analyzing sensory perception, this should be emphasized in the title, abstract, and introduction. The researcher should also explain the other methods used and the advantages of Spatial Syntax over them, mentioning that while it is a conventional method used in many studies, it is still relevant and effective for the current research.

Sixth, some statements in the paper seem to be results taken from other studies and should be carefully reviewed and verified for reliability, such as: “The integration of data technology can effectively improve residents’ quality of life, promote sustainability, and improve public services by integrating and highly integrating information about space.” Another example is, “In the study of spatial quality, most scholars only quantify the space and compare the data without further applying the quantification structure. A single quantification method and evaluation standard are bound to have limitations.” The conclusion drawn from these statements, which the researcher will build upon to complete their work, must be mentioned.

Seventh, the case study needs more depth and should be supported by evidence rather than personal or individual perceptions. The choice of the case needs to be more convincing.

Eighth, the use of the OSM (OpenStreetMap) online map collaboration program to obtain more complete street network data should be mentioned in the abstract, introduction, and beginning of the methods section. Highlighting the connection between the solution to the problem and the method will add novelty to this research and its unique applications.

This paragraph needs clarification: “This study used the literature analysis method to select indicators for multidimensional sensory perception.” It would be better to mention this earlier in the methods section.

Tenth, I had difficulty following the discussion results due to the paper’s lack of a specific structure, the confusion of methods, and the lack of clarity.

Finally, the conclusion section must be more robust and adhere to the scientific method. It should summarize the problem, the solution, the method used in future studies, and their importance to the international scientific community. The results in the conclusion should be helpful both practically and theoretically.

Notably, while the research demonstrates effort in exploring the spatial quality of urban underground pedestrian streets, concerns arise regarding the selection and justification of the cited works. The current reference list, heavily weighted toward scholars from a single country, needs a balanced representation of global research on this topic. Additionally, the criteria used to select these particular papers require further clarification. To enhance the research’s credibility and comprehensiveness, the authors should provide a more transparent explanation for their selection of references, ensuring a diverse range of perspectives and relevant international research are incorporated. This comment will strengthen the paper’s scholarly foundation and allow for a more robust topic analysis.

Author Response

我们根据您的建议对文章进行了全面修改，并回答了您提出的所有问题，详见文档：

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

 

By combining spatial syntax, multidimensional sensory perception and machine learning algorithms, this study proposes a multi-model, multi-indicator, multi-level spatial quality evaluation model, which is innovative in existing research on urban underground spaces. The study conducted field research in the underground pedestrian streets in Yangpu District, Wujiaochang, Shanghai, and collected a large amount of first-hand data, which provided a solid data foundation for the study. However, this paper still has many shortcomings and needs a lot of revision.

First, the literature review part can be more in-depth, compare more domestic and international related studies, and clarify the innovative points and shortcomings of this study based on existing research. The authors have only partially reviewed the literature on street space, and there are many evaluation indexes and methods for street space that the authors have not systematically reviewed, and in addition, there are many evaluations of indoor and underground spaces that also need to be included.

Secondly, the case selected in this paper is only the underground pedestrian street under TOD orientation, there are many types of underground pedestrian space, the evaluation of this case can not fully represent the underground pedestrian space, the authors should elaborate and adjust the title accordingly, otherwise it will cause ambiguity.

Third, the number of questionnaires of the authors is small, only 213 questionnaires in the case of 21 indicators. The credibility of the questionnaire results is questionable. In addition, the authors did not elaborate on how the control variables for the underlying information in the questionnaire were carried out, which also affects the credibility of the results.

Fourth, the drawings were drawn at a low level of legibility. It is recommended that the presentation of the graphs be optimized to improve the clarity and readability of the graphs.

Fifth, in the selection of multidimensional sensory perception indicators, there are more indicators in the visual dimension and fewer indicators in other sensory dimensions. The data collection and questionnaire distribution in the study were short and lacked periodic testing. Seasonal changes have a significant impact on the evaluation of underground pedestrian streets.

Sixth, the analysis of the spatial syntax and the evaluation of the questionnaire were cut off, and there were no obvious conclusions based on the spatial syntax reflected in the significant findings.

In conclusion, the thesis needs a lot of revisions.

 

Author Response

我们根据您的建议对文章进行了全面修改，并回答了您提出的所有问题，详见文档：

Author Response File: Author Response.docx