Statistical Learning Improves Classification of Limestone Provenance
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsProvenance of limestone in archaeology does pose a challenge, especially in the case of absence of any sedimentary structures or microfossils, as the authors correctly stated. The authors bring methodology to help resolve this issue. They clearly explain the methodology and analytical and statistical techniques they use to try to find the provenance of archaeological stone products. This methodology can be of assistance to other authors facing the same issues in their research.
I only suggest changing the title of the paper to "Statistical learning as a tool for improving classification of limestone provenance", but it is up to the authors.
Author Response
Thank you for your time and effort invested in reviewing and improving our manuscript.
Comments 1: I only suggest changing the title of the paper to “Statistical learning as a tool for improving classification of limestone provenance”, but it is up to the authors.
Response 1: After considering your suggestion, we decided to keep the shorter and simpler title. The authors think that current title is acceptable – It conveys the same meaning but is simpler and shorter. According to modern standards of scientific writing, shorter titles are preferred, furthermore statistics show that articles with shorter titles receive more citations.
Reviewer 2 Report
Comments and Suggestions for AuthorsA very interesting paper with future application possibilities. Only minor revisions are suggested.
Title: Please make the title more neutral. For example: The application of Statistical Learning for improving the classification of limestone provenance
Abstract: Please explain, possibly in brackets, what the Ig area (regio X) is and where it is located.
An English review from a native speaker will benefit the paper.
Line 69: Please indicate other possible materials that the proposed application can be used for (for example, other stones, ceramics?).
Figure 1: An additional map of Slovenia's location (and the site) will help readers unfamiliar with the European region.
Line 108: Could you please explain the protocol so that other researchers could use it in the future?
Line 113: What was the protocol for choosing these 53 samples?
Line 117: Why were 15 samples of the 53 used for the full analysis?
Section 4.2: Could the authors elaborate on whether their feature selection (especially the importance of Sr isotopes) would hold for different carbonate platforms, or whether the method must be recalibrated case by case?
Overall, while the geological detail is comprehensive, the archaeological interpretation could be strengthened. For example, what do these provenance results imply about Roman quarrying strategies, transport logistics, or economic networks in the Ig area? The addition of a relative bibliography (or a short commentary) will attract a broader set of readers.
Conclusions: I would suggest that the authors are preparing the conclusions in a proper paragraph and not using bullets.
Comments on the Quality of English LanguageThe paper would benefit from a native speaker's English review.
Author Response
Thank you for your comments that helped us to improve the manuscript.
Comment 1: Is the content succinctly described and contextualized with respect to previous and present theoretical background and empirical research (if applicable) on the topic? - ''Can be improved''
Response 1: We hope that the minor changes made throughout the manuscript made the paper more integrated and contextualized.
Comment 2: Title: Please make the title more neutral. For example: The application of Statistical Learning for improving the classification of limestone provenance
Response 2: The authors think that current title is acceptable – It conveys the same meaning, but is simpler and shorter. According to modern standards of scientific writing shorter titles are preferred, furthermore statistics show that articles with shorter titles receive more citations.
Comment 3: Abstract: Please explain, possibly in brackets, what the Ig area (regio X) is and where it is located.
Response 3: We included the required explanation in the abstract and also in the changed Figure 1.
Comment 4: An English review from a native speaker will benefit the paper.
Response 4: English review will be done by MDPI editing service.
Comment 5: Line 69: Please indicate other possible materials that the proposed application can be used for (for example, other stones, ceramics?).
Response 5: We included the proposed other materials that can benefit from the use of statistical learning for provenance determination.
Comment 6: Figure 1: An additional map of Slovenia's location (and the site) will help readers unfamiliar with the European region.
Response 6: Figure 1 was adapted in line with your comment.
Comment 7: Line 108: Could you please explain the protocol so that other researchers could use it in the future?
Response 7: We further explained the sampling protocol.
Comment 8: Line 113: What was the protocol for choosing these 53 samples?
Response 8: We detailed the protocol for choosing the archaeological samples.
Comment 9: Line 117: Why were 15 samples of the 53 used for the full analysis?
Response 9: We also prefer that more than 15 archaeological samples would be analyzed, however here this was not possible due to the budget limit.
Comment 10: Section 4.2: Could the authors elaborate on whether their feature selection (especially the importance of Sr isotopes) would hold for different carbonate platforms, or whether the method must be recalibrated case by case?
Response 10: An elaboration on Sr isotopes usefulness in other geological settings was added in section 4.4.
Comment 11: Overall, while the geological detail is comprehensive, the archaeological interpretation could be strengthened. For example, what do these provenance results imply about Roman quarrying strategies, transport logistics, or economic networks in the Ig area? The addition of a relative bibliography (or a short commentary) will attract a broader set of readers.
Response 11: A paragraph about the Roman quarrying strategies was added in the end section of the discussion.
Comment 12: Conclusions: I would suggest that the authors are preparing the conclusions in a proper paragraph and not using bullets.
Response 12: Conclusions were rewritten to avoid bullet points as you suggested.
Reviewer 3 Report
Comments and Suggestions for AuthorsVery interesting article; Some sugestiones are providedIf the authors are able to improve the manuscript.
All the statistical method is well explained and developed, but maybe could be more accesible for scholars not familiar with mathematics.
On the other hand, if you can contrast the data with XRD, MOP and SEM-EDX test your paper will be better and more citable.
Author Response
Thank you for your time and effort invested in reviewing and improving our manuscript.
Comment 1 :All the statistical method is well explained and developed, but maybe could be more accessible for scholars not familiar with mathematics.
Response 1: Our writing was initially tailored to the intended audience—geologists and archaeologists—so even the previous version of the manuscript already contained less technical and much more accessible explanations than e.g. Wikipedia, to suit the readers who are not statisticians or mathematicians. In version 2, minor corrections were made that will further ease the understanding. On the other hand, care had to be taken to keep the explanation statistically correct.
Comment 2: On the other hand, if you can contrast the data with XRD, MOP and SEM-EDX test your paper will be better and more citable.
Response 2: While we are currently unable to supplement our analysis with additional data, we have added a paragraph discussing the potential inclusion of other classification variables (e.g., XRD or SEM-EDX). Minor changes that contribute to the clearness of the conclusions were made throughout the manuscript.
