Extracting Features from Satellite Imagery to Understand the Size and Scale of Housing Sub-Markets in Madrid

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

An impressive paper given that this seems to be a new area of research for the authors judging by the reference list (apart from one paper co-authored by Arribas-Bel). The research on which the paper is based has been prosecuted with great thoroughness and the paper is tightly written, with suitable illustrative material. The only fault I find with it relates to sentence construction plus a few typos, see next box.

Comments on the Quality of English Language

As regards the few typos, a careful check would be sufficient, maybe even just at the copy-editing stage, starting with the spelling of ‘satellite’ in the paper’s title.

As regards sentence construction, there are more issues to deal with than I choose to list, but basically there are several places where two sentences which should be read separately are merely separated by a comma, not a full stop or even a colon. Just one example perhaps from page 2: line 57 ‘..over time, so far…’. Also some places where what is written is not a proper sentence, e.g. page 2 line 75 ‘For example, using k-means [7, 15, 26].’

A couple of slightly more substantive queries on page 8. Line 207 mentions ‘3x3 pixels’ – is this correct, given that this configuration is not mentioned elsewhere in the paper?

In line 220, what is meant by ‘..uses the features embedding to cluster patches into groups’?

Also, page 13, should the caption of Figure 6 read with ‘CH’ rather than ‘CB’ as it refers to Calinski Harabasz? But preferably spell out the full name.

Page 16, lines 475-6: should one of the two mentions of ‘internal’ be ‘external’?

And when discussing the later results, does ‘cluster labels’ have a very precise meaning rather than just referring to separate clusters (I would have expected ‘labels’ to include names or some such identifier/separator).

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

I am grateful for the opportunity to review this manuscript. Upon careful examination, I offer the following suggestions for its enhancement:

1.While the application of unsupervised machine learning to identify housing sub-markets is commendable, further elaboration on how this methodology advances beyond current models used in urban studies would strengthen the paper. Specifically, contrasting this approach with traditional statistical methods could highlight the innovation more clearly.

2.The use of MOSAIKS for feature extraction is an interesting choice. Expanding on why this method was chosen over others, particularly in terms of its effectiveness in capturing relevant features for housing market analysis, would be valuable.

3.Although external data from Idealista is used for verification, incorporating additional external validation methods, such as expert assessments or comparisons to governmental housing data, could further strengthen the findings.

4.A deeper dive into the methodological limitations, such as the potential for overfitting in the clustering process or the sensitivity of your model to hyperparameter settings, would provide a more critical analysis of your approach.

5.Detail any technical challenges encountered during the study, such as computational limitations, data integration issues, or challenges in applying the MOSAIKS framework to your specific dataset. Discussing how these were addressed or remain as limitations.

6.Illustration quality needs to be improved, the map's compass, scale, legend, and other elements cannot be missing.

Comments on the Quality of English Language

The level of English in the manuscript is generally good, with only negligible errors that do not compromise the clarity of the presentation.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have made some efforts to improve the manuscript quality, but I believe that the manuscript still has the following issues that need to be addressed:

1.Figures 1 and 2 require optimization in their graphic expression.

2.The flowchart in Figure 3 lacks clarity and should be reorganized and adjusted.

3.Consider the placement of legends in Figures 4, 5, 7, Supplementary Figure 1, and Supplementary Figure 2.

4.The expression of the north arrows and legends in Figures 8 and 9 needs to be optimized.

The figures and layout of the article do not meet the publication requirements, and it is recommended to carefully create and arrange the figures and layout.

Author Response

Please see the attachment

Author Response File: Author Response.docx