University Research for the Improvement of SDGs: A Framework for Mapping and Assessing SDG Science at the Country Level

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Final recommendation: Accept with minor revisions.

1. This topic selection holds certain value, especially as the current global education reform has entered the era of fourth-generation university competition. How to promote the sustainable development of universities is an important issue.
2. From the abstract, the structure appears complete. This article only considers the situation of universities in Brazil. If it is limited to one country, it is recommended to increase the sample size of selected universities. It is further recommended to conduct an international comparison, if possible. This suggestion is for the author to consider and choose.
3. The literature review section is relatively well-chosen. However, it only covers two aspects. I suggest adding some academic research, especially on the progress of indicator-related research, that is, the research on indicators related to sustainable development in universities.
4. The research methodology section feels odd. It seems more like a quantitative analysis of literature retrieval rather than the research we hoped to see on the sustainable development goals of universities. If it is limited to quantitative analysis of literature, it is not impossible.
5. Data analysis: It is recommended to attach the original data tables when publishing papers.
6. The conclusion section is basically normal.
7. In the references section, there is an overuse of journals such as "Journal of Cleaner Production" and "Sustainability". Given that this is a research topic related to universities, it is recommended to include some of the most crucial journals in the field of education.

Author Response

Dear Reviewer, we thank you for your valuable comments and suggestions to improve this article. I strive to implement all of them within the scope of the research and the journal's required time. Below, we present the answers to each suggestion and the actions taken to address them.

Reviewer 1:

Comment R1.1:

This topic selection holds certain value, especially as the current global education reform has entered the era of fourth-generation university competition. How to promote the sustainable development of universities is an important issue.

Response R1.1:

Thank you for the valuable comment. It shows that we are working in a valuable field: sustainability, education, and research in universities.

Comment R1.2:

From the abstract, the structure appears complete. This article only considers the situation of universities in Brazil. If it is limited to one country, it is recommended to increase the sample size of selected universities. It is further recommended to conduct an international comparison, if possible. This suggestion is for the author to consider and choose.

Response R.1.2: With agree with your comments. In fact, it is a limitation of the study. But due to the scope of the research and the time of execution, as established by the Sustainability Editorial by 23 February, we could not insert this new content in the article. But we will mention this limitation in the conclusions. As such, it is important to mention that in Brazil, we searched five most representative research institutions on SDG 2, which represent more 12,396 publications in a total of 23005 for the whole country.

Comment R1.3:

The literature review section is relatively well-chosen. However, it only covers two aspects. I suggest adding some academic research, especially on the progress of indicator-related research, that is, the research on indicators related to sustainable development in universities.

Response R1.3:

We added subsection 2.3 to the literature review, which presents the main recent contributions to the assessment of universities' performance on the SDGs.

 

Comment R1.4:

The research methodology section feels odd. It seems more like a quantitative analysis of literature retrieval rather than the research we hoped to see on the sustainable development goals of universities. If it is limited to quantitative analysis of literature, it is not impossible.

Response R1.4:

We added, at the beginning of the methodology, an argument that better justifies studying the university/research institute's scientific production related to the SDGs as an effective way to map and develop knowledge and technologies to improve the SDGs.

Comment R1.5:

Data analysis: It is recommended to attach the original data tables when publishing papers.

Response R1.5: This study involved the generation of dozens of tables (artefacts) from SciVal. As such, we can open the tables for readers on request.

Comment R1.6:

The conclusion section is basically normal.

Response R1.6: We cordially accept the comment.

Comment R1.7:

In the references section, there is an overuse of journals such as "Journal of Cleaner Production" and "Sustainability". Given that this is a research topic related to universities, it is recommended to include some of the most influential journals in education.

Response R1.7:

We researched and cited additional articles from the International Journal of Higher Education, a journal that publishes research on higher education. These articles approach SDGs or sustainability topics in the university context.

 

Reviewer 2 Report

Comments and Suggestions for Authors

The present manuscript proposes a three-level framework (subject areas, cluster names, and knowledge categories) to map and assess university research contributions to the Sustainable Development Goals (SDGs) at the country level. The framework is validated using SciVal data on SDG 2 (Zero Hunger) research across five leading Brazilian institutions between 2015 and 2024.

Overall, the article addresses a relevant and timely topic, particularly given the increasing pressure on universities to demonstrate measurable contributions to the 2030 Agenda. The paper’s main strength lies in its attempt to bridge macro-level bibliometric indicators and highly granular topic classifications through an intermediate “knowledge category” layer.

While the framework of the research is clearly described, its theoretical grounding is somewhat limited. The manuscript could benefit from several additions:

A clearer articulation of how this framework differs fundamentally from standard bibliometric mapping approaches.

Stronger theoretical anchoring in research evaluation literature or science policy theory.

A deeper explanation of why the “intermediate level” is conceptually necessary beyond practical convenience.

Currently, the contribution is more methodological than conceptual; this should be made explicit and better justified.

The study explicitly acknowledges (in the limitations section) that publication quantity is used as a proxy for research performance. However, this limitation is substantial. I consider there are some missing dimensions, including: citation impact, field-weighted citation impact, journal quality, collaboration networks and societal impact indicators.

Given that SDG research is inherently impact-driven, reliance solely on volume may oversimplify institutional contributions.

The grouping of cluster names into “knowledge categories” is central to the paper. However the criteria for grouping are not fully detailed. It is unclear whether the categorization was manual, automated, or based on predefined taxonomies. As far as I have noticed, no inter-coder reliability or validation process is reported.

Because this step involves interpretive decisions, methodological transparency is essential for replicability.

My recommenations are: strengthen the theoretical justification of the intermediate framework layer, clarify methodological procedures for knowledge category construction, regarding the study, please try, at minimum, to include one additional performance metric (e.g., citation-based indicator) to strengthen robustness, improve language precision and formatting consistency.

Author Response

Dear Reviewer, we thank you for your valuable comments and suggestions to improve this article. I strive to implement all of them within the scope of the research and the journal's required time. Below, we present the answers to each suggestion and the actions taken to address them.

Reviewer 2:

Comment R2.1:

The present manuscript proposes a three-level framework (subject areas, cluster names, and knowledge categories) to map and assess university research contributions to the Sustainable Development Goals (SDGs) at the country level. The framework is validated using SciVal data on SDG 2 (Zero Hunger) research across five leading Brazilian institutions between 2015 and 2024.

Response R2.1: We agree with the comment.

Comment R2.2:

Overall, the article addresses a relevant and timely topic, particularly given the increasing pressure on universities to demonstrate measurable contributions to the 2030 Agenda. The paper’s main strength lies in its attempt to bridge macro-level bibliometric indicators and highly granular topic classifications through an intermediate “knowledge category” layer.

Response R2.2: We are happy with the comment, which shows us the importance of the study and details its contribution to the literature.

 

Comment R2.3:

While the research framework is clearly described, its theoretical grounding is somewhat limited. The manuscript could benefit from several additions: A clearer articulation of how this framework differs fundamentally from standard bibliometric mapping approaches. Stronger theoretical anchoring in research evaluation literature or science policy theory.

 

Response R2.3:

In this new version, we consider that, as well as in bibliometric studies, this present study is based on articles' metadata, but in this case, centrality is based on statistics and qualitative analysis. The main difference of this study from current bibliometric studies is that it provides a deep qualitative perspective on the specific content studied in SDG 2 research. This model is applied to Brazilian universities, and can be replicated other similar situations.  

Comment R2.4:

A deeper explanation of why the “intermediate level” is conceptually necessary beyond practical convenience.

Response R2.4:

Proposing knowledge categories as an intermediary level between subject areas and cluster names serves as an integrative multilevel framework that enables a tiered assessment of the topics addressed by a research institution from a top-down perspective.  As a result, this model allows navigation through different scope levels of research institutions within an SDG.

 

Comment R2.5:

Currently, the contribution is more methodological than conceptual; this should be made explicit and better justified.

Response R2.5:

We seek to justify this further in the Discussion and Conclusion sections, arguing for the ease of use of the framework and its efficiency in mapping SDG science within research institutions at the country level.

 

Comment R 2.6:

The study explicitly acknowledges (in the limitations section) that publication quantity is used as a proxy for research performance. However, this limitation is substantial. I consider there are some missing dimensions, including: citation impact, field-weighted citation impact, journal quality, collaboration networks and societal impact indicators. Given that SDG research is inherently impact-driven, reliance solely on volume may oversimplify institutional contributions.

 

Response R2.6:

We agree with this consideration. In the “Discussion”, we note this limitation and consider the possibility of using other indicators to provide the best view of research institutions' contributions to advancing the SDGs. In fact, even though the quantity of publications can be an indicator of the contribution of a research institution to SDGs improvement, there are not enough to provide a comprehensive view of their contribution to the improvement of SDGs. Eventhoug in the field of bibliometric studies, other indicators such as, number of citations, cite score of journals where the articles are published, and field weight citation index can improve this view, it is possible to consider other indicators of production and transfer of knowledge, such as, patents, its licensing, and projects of technology transfer performed by the researcher institutions. It is cited as a limitation of the study.

 

Comment R2.7:

The grouping of cluster names into “knowledge categories” is central to the paper. However the criteria for grouping are not fully detailed. It is unclear whether the categorization was manual, automated, or based on predefined taxonomies. As far as I have noticed, no inter-coder reliability or validation process is reported. Because this step involves interpretive decisions, methodological transparency is essential for replicability.

Response R2.7:

We detailed in 3.1. subsection the activities performed for the obtention of the knowledge categories.

Comment R2.8:

My recommendations are: strengthen the theoretical justification of the intermediate framework layer, clarify methodological procedures for knowledge category construction, regarding the study, please try, at minimum, to include one additional performance metric (e.g., citation-based indicator) to strengthen robustness, improve language precision and formatting consistency.

Response R2.8:

We have endeavored to accomplish, within the possibilities of the current paper, all of these recommendations. We thank you for the valuable comments which contributed to the improvement of this article.

 

Reviewer 3 Report

Comments and Suggestions for Authors

Please see the attachment

Comments for author File: Comments.pdf

Author Response

Dear Reviewer, we thank you for your valuable comments and suggestions to improve this article. I strive to implement all of them within the scope of the research and the journal's required time. Below, we present the answers to each suggestion and the actions taken to address them.

Reviewer 3:

Review Report on the manuscript entitled

«University Research for the Improvement of SDGs: A Framework for Mapping and Assessing SDGs Research at the Country Level»

 

Comment R3.1:

The manuscript submitted for review focuses on mapping knowledge related to research on the Sustainable Development Goals at the national level, using the case of five Brazilian research institutions. The proposed model is validated through an empirical study based on SciVal-Elsevier data, covering five Brazilian universities in the context of research on Sustainable Development Goal 2, Zero Hunger, over the period 2015-2024.

Response R3.1: We agree with the comment.

Comment R3.2:

Overall, the manuscript is well structured and prepared at an appropriate scientific level, considering the specific nature of the study. At the same time, several conceptual and methodological aspects would benefit from further refinement. This report presents the results of a comprehensive assessment of the manuscript’s main sections. The Abstract is generally clear and well structured. It logically outlines the relevance of the study, the applied methodology, and the potential areas of practical application of the proposed approach. However, the abstract does not explicitly state the objective of the article and does not sufficiently highlight the scientific conclusion.  Given the descriptive and cartographic nature of the study, which does not aim at deep theoretical interpretation, a clearer formulation of the objective (specifically as mapping and systematization of research) would help to better manage reader expectations and reduce demands for interpretative depth.

Response R3.2:

We changed the presentation of the objective, stating, “the goals of this article are to introduce…” in the sense of stating clearly the article's purpose in relation to mapping the current performance of universities and research institutes in SDG science.

Comment R3.3:

In addition, a brief statement of the main scientific conclusion in the abstract is an important and essentially mandatory element, as it allows the reader to clearly understand the key intellectual contribution of the work.

Response R3.3:

We added a conclusive statement about the article’s contribution.

Comments R3.4:

The Keywords are generally appropriate and consistent with the topic of the manuscript. However, they could be further refined by explicitly indicating the research object, the methodological approach, and the country of analysis, which would improve their informativeness for readers and search engines.

Response R3.4:

We changed some keywords to better specify the articles’ content.

Comments R3.5:

The Introduction deserves a positive evaluation for its conciseness and logical flow. The authors consistently guide the reader from the global challenges of SDG implementation to the role of universities and research institutions in advancing the SDGs, and further to the rationale for mapping research in the field of sustainable development. The structure is coherent and clear, and the transitions between analytical levels are smooth and well justified.

Response R3.5:

 Thank you for the observations. They make us perceive that our efforts are in the right direction.

 

Comment R3.6:

 The Literature Review section is well structured and substantive. It presents a broad overview of scientific publications published between 2015 and 2025. Particularly strong is the coverage of studies addressing the role of universities and research institutions in the implementation and study of the SDGs, with an emphasis on educational, strategic, and research dimensions. The review logically leads to the chosen research approach and clearly justifies the need for mapping SDG-related research.

 

Response R3.6: 

Thank you for the observations. They make us perceive that our efforts are in the right direction.

 

Comment R3.7:

At the same time, if the authors consider it appropriate, this section could be slightly strengthened by adding a short contextual paragraph on the specific features of SDG implementation and the scientific discourse in Brazil. Such a paragraph would serve an orienting function and help readers better understand the national context of the subsequent empirical analysis.

Response R3.7:

We added at the end of section 2.2. a paragraph presenting researches in Brazil that addresses SDGs in the university context.  

 

Comment R3.8:

 Alternatively, placing this contextualization in the methodology section also appears justified. The Methods section is detailed and technically sound. The authors clearly emphasize the use of SciVal data and consistently describe the three stages of the mapping process: identifying Brazil’s position in terms of SDG research performance and selecting the most researched SDG in the country (SDG 2, Zero Hunger); identifying Brazilian research institutions with the largest scholarly output on SDG 2 (selection of five institutions out of 200); analysing the knowledge areas in which SDG 2– related publications are produced within these institutions.

Response 3.8:

We agree with this comment.

Comment 3.9:

The chosen methodological toolkit is generally consistent with the research objectives and with established practices in empirical studies in this field. Importantly, the authors apply two complementary analytical approaches: an internal approach (assessing the relative importance of knowledge categories within each institution) and an external approach (aimed at identifying differences between research institutions). Given this methodologically sound and appropriate approach, it may be useful to complement the section with a brief informational overview of the selected institutions, both in general terms and in relation to SDG 2. This would allow readers to better trace potential patterns and relationships and to more clearly interpret the results.

 

Response 3.9:

We complemented the description of the research object by presenting the importance of the five chosen institutions, given their high representativeness in SDG 2 research in the Brazilian context. The selection of the five institutions in this study, a research institute and four universities, is due to the fact that they are the best-ranked institutions in terms of publications related to SDG 2 in Brazil. Together, these five institutions account for 12396 publications related to SDG2, out of a total of 23008 publications, representing 53,87% of the total publications of Brazil on this SDG, according to SciVal.

 

 

Comment R3.10:

The Results section presents the findings in a clear and well-organized manner, in line with the chosen methodology. The main focus is on the quantitative comparison of research profiles of institutions within the context of SDG 2. At the same time, the interpretative dimension of the results is limited, which corresponds to the descriptive and cartographic nature of the study.

Response R3.10:

Due to the size limitations of the article, we opted to improve the interpretations in the “5.Discussion”.

 

Comment R3.11:

The tables are correctly constructed; however, visualizing part of the results could further enhance clarity and readability.

Response R3.11:

We effort to better explain the content of tables, improving their internal description in some situations.

 

 

 

Comment R3.12:

The Discussion section is consistent with the overall logic of the study and appropriately relates the findings to existing research. Nevertheless, the discussion remains largely descriptive and focuses mainly on confirming the relevance of the proposed approach. In the reviewer’s opinion, this section could be strengthened by a clearer interpretation of the results and by a more explicit discussion of their implications for research management and science policy in the SDG context, which would help to further “individualize” the proposed methodology.

Response R3.12:

We sought to improve the interpretation of results in the Discussions and added a paragraph at the end that discusses the implications of the framework for public agents in establishing policies linking SDG science to communities' public demands, with the purpose of effectively increasing municipalities' and regions' SDG performance.

 

Comment R3.13:

The Conclusions section is generally consistent with the results of the analysis and summarizes the main points of the preceding sections. The authors appropriately identify the main contribution of the article as the development of a replicable framework for mapping SDG-related research at the country level, as well as the creation of a research map for SDG 2.

Response R3.13:

 We agree with this comment.

Comment R3.14:

A positive aspect is that the authors clearly outline the limitations of the study and suggest possible ways to address them. At the same time, to improve clarity and usability, it could be useful to summarize the strengths and opportunities, as well as the limitations and constraints of the proposed model in a separate table. This would allow readers to more easily navigate the framework’s potential and methodological boundaries.

Response R3.14:

In this new version, we detailed the strengths and limitations of the research in the Conclusion.

 

Comment R3.15:

In addition, given that the creation of a research map is identified as one of the key contributions, the conclusions could place greater emphasis on the mapping results themselves.

Response R3.15:

We put more emphasis on the mapping results in the Discussion, whereas in the Conclusion we prioritised the ontological contribution of the models, that is, beyond the specific empirical data obtained from SciVal.

 

Comment R3.16:

In particular, the authors could explicitly summarize whether the identified research specialization of institutions corresponds to the knowledge areas that dominate their publication profiles, and what similarities or differences between institutions were observed.

Response R3.16:

This was performed in the additional description of the universities in “3.Method”.

 

Comment R3.17:

As a direction for future research (not as a critique but as a constructive suggestion), the study could be extended by introducing a dynamic, time-based analysis of research mapping. Given the ten year period covered, future work could examine changes in research orientations at both the institutional and national levels, assess whether research priorities have been consistent or shifting over time, and explore the extent to which such dynamics may reflect the effectiveness of science policy in the SDG domain. The Reference list includes 61 sources, predominantly published within the last ten years, and is fully consistent with the topic of the study. Overall, the manuscript represents a sound and well-structured scientific study aimed at mapping and systematizing research related to the Sustainable Development Goals. The applied analytical methods are in line with current scientific practice, and the results demonstrate both analytical and practical potential. The reviewer’s comments are intended as recommendations and are aimed at further strengthening the scientific interpretation of the findings.

Response R3.17:

We agree and thank you for the valuable comments. We strive to improve this article version based on all the suggestions.  

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The revised manuscript presents a clearer and more structured version of the proposed three-level framework (subject areas – knowledge categories – cluster names) for mapping SDG science at the country level. The methodology is now more explicitly described (Section 3), and the creation of the intermediate “knowledge categories” layer is better operationalized than in the previous version.

The article continues to offer a methodological contribution rather than a theoretical one, and this positioning is now more explicit in the Discussion and Conclusions. The empirical application to SDG 2 in Brazil remains comprehensive and detailed.

There are still some issues with the article.

1. Although improved, the theoretical foundation remains relatively weak. What remains underdeveloped: clear positioning within science policy evaluation theory, conceptual justification of why “knowledge categories” constitute a theoretical innovation rather than a practical grouping strategy, engagement with literature on research specialization, institutional differentiation, or innovation systems

My recommendation is to  add a subsection in the Discussion explicitly framing this contribution in: research evaluation literature, national innovation systems theory, strategic research management. Currently, the manuscript overemphasizes methodological novelty and underemphasizes conceptual contribution.

2. While improved, the grouping process remains partially subjective, as there are: no coding protocol is described, no validation procedure is reported and no example of how clusters were grouped step-by-step is provided. This remains the most vulnerable part of the framework. Without clearer operational rules, replicability may be questioned.

My recommendation is to include, if possible: an appendix with a sample grouping logic, or a methodological flowchart explaining classification criteria, or at least a clearer description of decision rules.

3. Although improved, the manuscript still contains: minor grammatical errors, inconsistent ordinal notation (“3,1th”, “3,1rd” - page ten - both top and bottom of the table), occasional awkward phrasing and incomplete formatting in references (e.g., missing entries, numbering gaps - mainly in the first part)

Author Response

Dear Reviewer,

Now we present the third version of the article, in the second review round. I would like to thank you for the additional suggestions, which, in fact, we believe contributed to more clarity and improved the article's contribution. We strive to accomplish all of them, and we are at your disposal for any additional adjustments if necessary.

Cordially,

Authors.

Comment 1:

Although improved, the theoretical foundation remains relatively weak. What remains underdeveloped: clear positioning within science policy evaluation theory, conceptual justification of why “knowledge categories” constitute a theoretical innovation rather than a practical grouping strategy, engagement with literature on research specialization, institutional differentiation, or innovation systems

My recommendation is to  add a subsection in the Discussion explicitly framing this contribution in: research evaluation literature, national innovation systems theory, strategic research management. Currently, the manuscript overemphasizes methodological novelty and underemphasizes conceptual contribution.

Response:

We have searched for additional literature on the topics above and have updated the end of section 2.3 in the literature review. In the discussion, we follow the above recommendations and examine the contribution of the model proposed in Section 5.2.

 

Comment 2:

While improved, the grouping process remains partially subjective, as there are: no coding protocol is described, no validation procedure is reported and no example of how clusters were grouped step-by-step is provided. This remains the most vulnerable part of the framework. Without clearer operational rules, replicability may be questioned.

My recommendation is to include, if possible: an appendix with a sample grouping logic, or a methodological flowchart explaining classification criteria, or at least a clearer description of decision rules.

 

Response:

We agree with your comment, which in fact addresses a critical issue in the paper that should be clarified. Below, we provide a detailed description of these aspects. Due to space reasons, we synthesized the description below and included it in Section 3.1 to better explain the data treatment and remaining methodological actions.

Step 1 – The generation of prior reports before the reports used in this research.

Before addressing the reports used directly in this research, it was necessary to generate from SciVal general reports about the current world and the Brazilian situation in the generation of scientific knowledge in the form of publications, indexed in Scopus. There were generated the following reports based on SciVal data:

Report 1 – The current world quantity of publications related to each SDG: There are 16 prescreens in this file, one for each of the 16 SDGs, accounted in the Scival data. From this, we generated a world ranking and verified that the best Brazil position was in SDG2 (zero hunger) and SDG15 (life on land), with the fifth world position in them, with the respective quantity of publications in these two SDGs, 23008 and 25997. Due to the prominence of Brazil in world food production, we decided to address this issue in the article.

Report 2 – The ranking of Brazilian research institutions best positioned in SDG2. In this step, we generated from SciVal a ranking of the quantity of publications in SDG2 of the Brazilian institutions. We verified that the first five institutions, respectively, Embrapa, USP, UNESP, UFV and UFLA, are responsible for 12,396 publications in a total of 23.008, corresponding to a percentage of 53,87%. 

 

Step 2 – The generation of reports from Scival about the quantity of publications by subject areas and the relation of publications to specific topics.

In this step, two report types were generated for each of the five Brazilian institutions, corresponding a total of 10 reports.

The first type of report:

A quantitative report was generated, presenting the number of publications by subject area, i.e., a general category used in the Scopus database to classify articles. For example, in the case of Embrapa, the first-ranked in Brazil for SDG2, 26 subject areas were presented, along with the respective number of publications related to them. The respective reports were used to construct Table 2, which ranks subject areas within each research institution by the number of publications.

               Even though the number of publications by subject area was an important metric, the reports generated were not detailed; that is, they did not associate each publication with its subject area but only presented the total number of publications for each subject area. Given the research scope, we need greater detail on the generation of the second type of report, as described below.

The second type of report:

               This type of report was generated with the list of all publications of each research institution in the SDG2, selecting several parameters in the SciVal whereas we emphasize the following: “title”, which corresponds to the article title; “All Science Journal Classification (ASJC) field name”, which is a classification in the Scopus systems, of the fields that the journal/conference where the publication was published; “topic cluster name”, corresponds to a content classification in the Scopus, where each publication is associated with only one topic cluster name.                  Departing from the data generated through the SciVal in the report, a second sheet was created, where the data from the first sheet was copied and subsequently ordered by the “topic cluster name”. After this, the quantity of articles in each topic cluster name group was counted. For example, in the case of Embrapa, 253 topic cluster names were identified, in a total of 3825 publications. In this case, the cluster name with the major quantity of publications was Soil Carbon Dynamics in Agricultural Ecosystems with a total of 457 publications, followed by Methane Reduction through Feed Supplementation Strategies with 188, and so on, until to reach topic clusters names related to only one publication.

               Through the ordering of the quantity of data of publications associated with topic cluster names, permitted the verification of considerable asymmetries in this relation, and high level of concentration of the quantity of publications in a small number of cluster names. As such, we generated Table 5, where we accounted for each research institution the quantity of publications associated with the 20 best cluster names in each institution. For example, in Embrapa case, the quantity of publications associated with the 20^th most frequent cluster name was 2125 publications, in a total of 3825 publications, covering 55,55% of the publications. Similar situations were verified in the data of the four remaining research institutions. This result permitted us to conclude that the large number of topic cluster names for each research institution required a mechanism for synthesis and classification. Considering the in the case of Embrapa, there was identified in 26 subject areas in the high level classification, and 253 in the bottom level, this situation lead us conclude the necessity of a intermediary categorization, capable of group the topic cluster names according to similarity, presenting at the same time a broad view of the specific issues addressed in the SDG2 research of each institutions, but with a minor quantity of categorial elements.

Step 3 - The identification of knowledge categories:

               The need to create a superior classification category based on the topic cluster name arose from the large number of topic cluster name categories identified in each research institution's report. In this stage, the researchers, through independent review, followed by a meeting for nomenclature unification and validation, associated the cluster names with knowledge categories. In the case of Embrapa, we created a total of 28 knowledge categories, whereas the first 20 covered 96,23% of the articles, which is considerably superior to the similar classification based on topic cluster name. As a result, even though the creation of the construct knowledge category permitted an intermediary element between the constructs' subject areas and topic cluster name, it also permitted a high level of mapping of the issues covered in the publications and a map with easy visualisation.

As such, we believe that the new argumentation in Section 3.1 better explains Figure 1, which presents a synthesis (a flowchart) of the main steps performed in the research.  

 

Comment 3:

3. Although improved, the manuscript still contains: minor grammatical errors, inconsistent ordinal notation (“3,1th”, “3,1rd” - page ten - both top and bottom of the table), occasional awkward phrasing and incomplete formatting in references (e.g., missing entries, numbering gaps - mainly in the first part)

Response: We conducted a new grammatical revision of the text.

Author Response File: Author Response.docx