EducaWood: a Semantic Web Application for Forestry Education

I would like to thank the authors for their efforts to significantly improve the quality of the manuscript. I recommend a minor revision based on the need to restructure the presentation of EducaWood application, and add more details about the requirements and related works comparison. I am confident that this manuscript is on track for acceptance following a successful revision.

(1) To further highlight the quality, importance, and impact of the described application, I would suggest:
- explaining who identified the requirements (Section 3.1), and how. It would be useful to report whether these are the minimum requirements and, if so, why. Furthermore, as the requirements seem to be a fundamental component of the application development, it would be insightful to include a table mapping the requirements to the ontology elements.
- adding, in Sections 2.1 to 2.4, a short paragraph or a few words after the presentation of the related works regarding their comparison with the current work. Also, I would argue that page 5, lines 37-43 do not belong in the related works section.

(2) Regarding the clarity and readability of the describing paper I would recommend:
- splitting Section 3, "Design of EducaWood", into two sections one about the design and data aspects (i.e. "Design of EducaWood"), and one about the usage of the application (i.e. "Usage of EducaWood"). Currently, the two elements are mixed, which hinders the clarity of the manuscript. Additionally, I would appreciate seeing a usage scenario of the application that corresponds to the learning objectives of forestry education of Section 1 page 1, lines 46-49 and page 2, lines 1-5.
- converting the website footnotes into bibliography items (i.e. @misc), especially the repeated ones.
- adding more explanation in the Table 1 description in the manuscript regarding the purposes of the chosen domains.

Minor comments:
- Fig. => Figure
- In Table 2, name the challenges instead of using #1 or #3
- In Section 3.1, page 7, lines 24-27, consider rephrasing the "should" phrases
- Teachers words should be in quotation marks, not just in italics
- I would suggest removing:
* the single-sentence paragraphs
* any text announcing that details will come to another section (i.e., page 10 lines 38-39), except for the last paragraph in introduction
* "Note that" and "take into account that"

Second Review for EducaWood: a Semantic Web Application for Forestry Education

Review #1
- Reviewer Concern: You are highlighting forestry education, but mention Tree Management. I would recommend spending some thoughts creating different personas of User Types and what would be their use cases. I would assume that also Analysis/Layer Topics (e.g., highlighting a big amount of dead trees) would be necessary then. You could select representatives of personas and do lists with user acceptance tests.
- My Comments on revision: Partially addressed. Future work acknowledges this concern, but these are not yet fully developed

- Reviewer Concern: User Requirements of Personas should be logically linked with your functional requirements, especially in connection with Challenge 3. You should provide a scientific method to evaluate the quality of the end result. Your current requirements are basically boolean (it works or it does not) or, in the case of low latency, not very specific. Therefore, it is not possible to really grasp the quality of the implementation
- My Comments on revision: Partially addressed. Authors tried to establish the local link in the writing, but still didn't provide a solid scientific evaluation method and justify it

- Reviewer Concern: Please define your test cases and design them in a way to evaluate the current boundaries of the application (e.g., generate a Million Trees randomly and see how it affects loading times, render times, visualization quality, and usability in different scenarios)
- My Comments on revision: Addressed. Authors included writing where they have tested with 1.4 million trees.

- Reviewer Concern: Same thing should go for the performance of the API Calls under stress. As you were highlighting, the integration of multiple sources is challenging, so you should provide some metrics and gradients depending on loads
- My Comments on revision: Addressed. Authors included latency numbers for the API calls.

- Reviewer Concern: In the SPA Mode,l I got confused that classes are named Annotations and subclasses of Annotations when I would personally interpret them as Classes. Especially since the Annotation Property is a part of Ontology Standards. That creates an ambiguity that I would avoid in the model
- My Comments on revision: Addressed. Authors included a footnote where they address the overload of the terms

Review #2
- Reviewer Concern: EducaWood work seems quite similar to Forest Explorer, which has been published in previous papers (see references 38, 41, 51 in the paper). However, the paper does not explain how it advances the contributions made in these papers
- My Comments on revision: Addressed. Authors added writing that presents what exactly is differentiable

- Reviewer Concern: The related works section does not include past works similar to EducaWood (papers 41 and 51), and only briefly describes previous work done in paper 38. Moreover, this section does not mention how the described literature is relevant to the paper's work, which limitations were imposed, and which parts this paper adapts, and advances compared to the state-of-the-art. Further, Section 2.2 is a mix of different elements (visualizations, REST APIs etc), making it difficult to follow and understand their connection to the paper contributions; I would recommend rewriting this section or splitting it into more sections, each with distinctive content
- My Comments on revision: Addressed. Authors rewrote the whole section into more subsections for clear transition and flow

- Reviewer Concern: As the paper focuses on the EducaWood interface, it lacks a clear connection and advancement to the semantic web, especially since the architecture of EducaWood and the creation of annotations seem very close to those of Forest Explorer. Consequentially, I would recommend clearly stating the differences of the current paper output to previously published works. Also, it would be helpful adding in Table 2 dependencies related to SW. Moreover, the paper can be benefited of a section describing the underlying ontology and knowledge graph in numbers, and comparing them with state-of-the-art resources and the user-interface connections they provide
- My Comments on revision: Partially addressed. The STA ontology is mentioned along with a few contributions they make with this paper, but there is limited to no quantitative comparison to other ontologies through dimensions like expressivity, reusability of known vocabularies, etc.

- Reviewer Concern: I am not sure I understand well Table 1, as I would expect having the namespaces used in the ontology; however, I am not certain this is the case, as no information is further provided in the text. Moreover, the namespaces used raise concerns about best practices for ontology development as the same namespace (http://educawood.gsic.uva.es) is reused under different prefixes.
- My Comments on revision: Authors do not address or justify the heavy reuse of the base namespace. It's not a major issue, but needs to be justified or addressed.

- Reviewer Concern: Which aspects of the system and interface are aligned with which pedagogical and learning settings? Currently, the paper's discussion of educational aspects takes place only at the end of the paper (Sections 4 and 5). This leads to plenty of unsupported claims about the applicability of EducaWood in different educational levels, learning objectives, learners' interests, interdisciplinary learning, collaborative learning, and logical awareness, because it misses any connection with the EducaWood system and interface
- My Comments on revision: Addressed. Authors extended Section 4.2 Pilot study and a new conclusion section, which clarifies and addresses the aspects better

- Reviewer Concern: Section 4.2, regarding the experiment, such as the students background knowledge checks (if any), students introduction to EducaWood steps, an example of the annotation session and verification, more details about the design and learning objectives of the experiment and the choice of different steps (min 20 trees, forestry management) motivating the usage of SUS questionnaire, what type of teachers feedback was provided, which was the alternative paper-pencil activity and how EducaWood is making thinks better/easier, qualitative analysis and presentation of the students annotations (in comparisson with their scores) and discussion of minor bugs and how they might have affected the UX
- My Comments on revision: Partially Addressed. Authors included details in the section, like SUS scores, qualitative feedback approach, and details of the approach

- Reviewer Concern: Regarding the educational contribution, the results of SUS (75% with s.d. 11.5) seem very similar to Forest Explorer SUS results (75% with s.d. 16 in general. If the two systems are similar (to which degree is left to the revision to be clarified), I would need to see how EducaWood is better since the SUS results do not demonstrate any potential for statistical significance
- My Comments on revision: Not addressed. No direct statistical or methodological comparison with Forest Explorer is made

- Reviewer Concern: Regarding the design of web applications, the paper claims to contribution to the "good practices", but it is unclear which are those and how they apply them in EducaWood
- My Comments on revision: Addressed. Authors discuss Map-based interaction, CRAFTS API abstraction, data loading with geospatial cells, REST + JSON, etc

- Reviewer Concern: Clarity and Readability
- My Comments on revision: Addressed in most places

Review #3
No major concerns in Review 3