A Spatial Case-Based Reasoning Method for Healthy City Assessment: A Case Study of Middle Layer Super Output Areas (MSOAs) in Birmingham, England
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe manuscript is dedicated to an interesting and relevant problem and corresponds to the theme of the journal. The work examines the idea of ​​using spatial data for Healthy City Assessment, which is quite obvious and valuable from both fundamental and applied points of view.
The proposed idea has been brought to the level of methodology, satisfactorily described, the results of its use have been confirmed and well substantiated. Strong point of research is using of basics methodologies - 1st Tobler's Law, Gestalt theory, etc.
Authors correctly highlighted in the Introduction that HIA is more declaration. There are lack of theory, empirics. Also we must keep in mind that current science have no sufficient definitions of health, life, disease, etc. From this point of view core approach for health cities must be empirical, therefore the general logic of the authors' reasoning looks quite valid. Adding a spatial dimension to the approach seems quite natural and reasonable.
Some remarks for possible improvement.
1. Fig. 1. What does the color differentiation of the three blocks in Fig. 1 mean? The output comes from the green block, but the arrows lead to the red block and there is not a single exit from red to green or orange. It is recommended to explain the purpose of these blocks and the information circulation scheme in it in the text. And correct if required.
2. Fig. 2. It is recommended to use a different projection that does not distort geography so much, generally improve the quality of maps and show the entire territory of the islands - for example, the whole of Ireland. Northern Ireland is definitely not an island, which is essential for discussing health issues. In general, the quality of the maps in the article is low, which complicates the perception of the material.
3. 2.2.1. Data Sources. We must keep in mind that ad hoc selection of data based on POI (tobacco, alcohol, street view etc.) could be biased due to lack of integrity. Of course no other data, and no theory behind... It is just a remark for future investigations.
Author Response
|
Comments 1: Fig. 1. What does the color differentiation of the three blocks in Fig. 1 mean? The output comes from the green block, but the arrows lead to the red block and there is not a single exit from red to green or orange. It is recommended to explain the purpose of these blocks and the information circulation scheme in it in the text. And correct if required. |
|
Response 1: Thank you very much for pointing out this issue. We deeply apologize for the confusion caused by the colors of the three areas in Figure 1. In fact, this is just a research framework diagram, and the original purpose of the three colors was to differentiate the processes of different parts: the green on the left represents the CBR process, the red in the middle represents the modeling process of HCSCBR, and the orange on the right represents the corresponding algorithms used in the modeling process. To avoid any misunderstanding for you and other readers, we have redrawn this figure and standardized the colors to make it more academic and readable, and have made the corresponding changes in the manuscript (lines 118-129). Additionally, since the framework diagram clearly expresses the content, we believe that no further explanation is needed in the text. If you have any further suggestions or comments, we are very willing to listen and make the necessary modifications. |
|
Comments 2: Fig. 2. It is recommended to use a different projection that does not distort geography so much, generally improve the quality of maps and show the entire territory of the islands - for example, the whole of Ireland. Northern Ireland is definitely not an island, which is essential for discussing health issues. In general, the quality of the maps in the article is low, which complicates the perception of the material. |
|
Response 2: Thank you very much for pointing out this issue. As per your suggestion, we have made the corresponding changes in the manuscript (line 152-163). |
|
Comments 3: 2.2.1. Data Sources. We must keep in mind that ad hoc selection of data based on POI (tobacco, alcohol, street view etc.) could be biased due to lack of integrity. Of course no other data, and no theory behind... It is just a remark for future investigations. Response 3: Thank you for your insightful reminder. In our future research, we will reduce potential biases that may arise from collecting temporary POI sectional data. |
|
|
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe theme is very interesting and the methods are clearly described. The section of discussion of missing. The structure of the paper is correct. The references are relevant with the examined theme. I believe that the paper can be published in its current form.
Author Response
|
Comments 1: The theme is very interesting and the methods are clearly described. The section of discussion of missing. The structure of the paper is correct. The references are relevant with the examined theme. I believe that the paper can be published in its current form. |
Response 1: Thank you very much for your high praise. We have added "3.6. Discussion Summary" to the manuscript.(line 580-594)
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe manuscript adopts a multi-source fine-grained dataset covering environmental determinants and health outcomes. Using the MSOA scale of Birmingham, England, as the spatial unit of geographic research, it proposes an innovative Health City Evaluation Spatial Case-Based Reasoning (HCSCBR) model. This method is not only applicable for health evaluation of small-scale units within a city but can also be extended to evaluation between city scales, demonstrating good innovation and commendable value. However, to further improve the quality of the manuscript, the authors need to clarify or make modifications to the following issues:
Line 43-44: “In urban planning, there are three main types of healthy city assessment methods: qualitative, quantitative, and composite.” It should be clarified which are qualitative, which are quantitative, and which are composite research methods or models, without mixing them together.
Line 80-82: “As a future development direction, the geographical environment is integrated as a set of spatial driving factors in case retrieval, amendment, and reuse computation, though its practical application poses challenges.” According to the focus of the study, it is suggested to state that “spatial case and case amendment” is the practical application challenge.
Line 80-82: “Spatial features enhance the accuracy and efficiency of case retrieval and reuse, addressing complex, heterogeneous geographic problems.” The manuscript mentions solving spatial heterogeneity issues; the authors should further clarify where these issues are addressed and suggest adding this content to the text if necessary.
Line 344-345: “Sk is the set of similar cases obtained according to a predefined threshold.” Please clarify what the threshold is and its function.
Line 425: Figure 3. Pre-organization of the case: (a) 29 cities in England generate 3 clusters; the figure is not clear enough, please revise it.
Line 466-469: “To meet the experimental requirements, all the above experiments were carried out on the Windows 11 operating system using Python implementations based on open-source libraries such as NumPy, pandas, SciPy, sklearn, and Keras.” It is recommended to supplement the hardware environment description, such as CPU and memory.
Line 577: 4. Conclusion. Add a subsection as “Discussion” before this section to further summarize and discuss the experimental results.
Comments on the Quality of English LanguageAll are OK but please pay attention to some minor mistakes
Author Response
|
Comments 1: Line 43-44: “In urban planning, there are three main types of healthy city assessment methods: qualitative, quantitative, and composite.” It should be clarified which are qualitative, which are quantitative, and which are composite research methods or models, without mixing them together. |
|
Response 1: Thank you very much for pointing out this issue. We have clarified which methods are qualitative, which are quantitative, and which are comprehensive in the manuscript (lines 44, 45, 47). |
|
Comments2: Line 80-82: “As a future development direction, the geographical environment is integrated as a set of spatial driving factors in case retrieval, amendment, and reuse computation, though its practical application poses challenges.” According to the focus of the study, it is suggested to state that “spatial case and case amendment” is the practical application challenge. |
|
Response 2: Thank you very much for pointing out this issue. We have made the corresponding changes in the original text (lines 83-84). |
|
Comments 3: Line 80-82: “Spatial features enhance the accuracy and efficiency of case retrieval and reuse, addressing complex, heterogeneous geographic problems.” The manuscript mentions solving spatial heterogeneity issues; the authors should further clarify where these issues are addressed and suggest adding this content to the text if necessary. |
|
Response 3: Thank you very much for pointing out this issue. Spatial heterogeneity is not the focus of this study. However, we have actually addressed the issue of spatial feature heterogeneity in "2.3. Case Pre-organization" by using spatial clustering of MSOA centroid points to pre-organize cases, thereby forming highly similar sub-cases to optimize retrieval efficiency and location accuracy. |
|
Comments 4: Line 344-345: “Sk is the set of similar cases obtained according to a predefined threshold.” Please clarify what the threshold is and its function. |
|
Response 4: Thank you very much for pointing out this issue. When retrieving similar cases, a similarity threshold is set, and only when the similarity between a case and the current problem exceeds this threshold will the case be selected. Due to our oversight, this value was omitted; it should be 0.85. In fact, this threshold is also mentioned in "Experiment Three: Real-World Application" where it is set to 0.85 (line 347). |
|
Comments 5: Line 425: Figure 3. Pre-organization of the case: (a) 29 cities in England generate 3 clusters; the figure is not clear enough, please revise it. |
|
Response 5: Thank you very much for pointing out this issue. We have revised Figure 3 and updated it in the manuscript, using different colors to distinguish between different clusters for clearer representation (line419-427). |
|
Comments 6: Line 466-469: “To meet the experimental requirements, all the above experiments were carried out on the Windows 11 operating system using Python implementations based on open-source libraries such as NumPy, pandas, SciPy, sklearn, and Keras.” It is recommended to supplement the hardware environment description, such as CPU and memory. |
|
Response 6: Thank you very much for pointing out this issue. We have added the following information to the manuscript: "The hardware environment used is as follows: CPU is 13th Gen Intel(R) Core(TM) i7-13700 with a frequency of 2.10 GHz, and the RAM is 32.0 GB." (lines 471-472). |
|
Comments 7: Line 577: 4. Conclusion. Add a subsection as “Discussion” before this section to further summarize and discuss the experimental results. |
Response 7: Thank you very much for pointing out this issue. We have added "3.6. Discussion Summary" to the manuscript.(line 580-594)
Author Response File: Author Response.pdf
