Overall comment: Although this study has an interesting approach to study Wikipedia knowledge graph from the scientometric perspective, it is super-long (32 pages), mixing in-depth theoretical arguments with the broad quantitative analysis of metadata indicators (article attributes). This could be problematic to understand the contrast being investigated. I think authors may focus on quantitative part by answering relevant RQs to metadata indicators (e.g., edits, length, views) and publish another expanded theoretical paper on the topic. Discussions about the associations between metadata indicators (or informetric indicators as defined by authors) could be much deeper and intelligible for readers (Figures 4 and 5), highlighting how Wikipedia article attributes are associated and how high-quality contents could be produced. Although authors have gathered big data from Wikipedia, it is still known which factors may influence Wikipedia article quality due to domain differences on the subject. I don’t think including one correlation matrix table showing associations between variables could be sufficient without extra arguments, linking results to the some previous studies. For example, do Wikipedia articles with more editor contributions tend to attract more readers (views)? It is not clear why there were no correlation between views with edits and editors (0.07) but moderate association found between views and talks (r=0.56), for example. An argument about factors affecting article quality could be interesting for this.
Wilkinson, D. M., & Huberman, B. A. (2007). Assessing the value of coooperation in wikipedia. arXiv preprint cs/0702140.
Blumenstock, J. E. (2008, April). Size matters: word count as a measure of quality on wikipedia. In Proceedings of the 17th international conference on World Wide Web (pp. 1095-1096).
Zhang, H., Ren, Y., & Kraut, R. E. (2020, January). Mining and Predicting Temporal Patterns in the Quality Evolution of Wikipedia Articles. In HICSS (pp. 1-10).
Other comments

Table 1. and page 6: The review of literature is rather enumerative and could include some information about data sets and outcomes (e.g., correlations between variables). For instance, one study cited in Table 1 extracted 4 million citations to publications (e.g., DOI, PMC, PMID, and ISBN) (Singh et al., 2020) and another investigation analysed over 1.9 Wikipedia articles and over 824,000 references to scientific articles (Nicholson et al., 2021) and hence some statements such as “previous research around Wikipedia information are mostly rather ad hoc case studies, lacking large-scale and comprehensive perspectives on the study of the platform” are arguable. I don’t think Table 1 is informative enough to be included with the current contents.
Table 1. What do you mean by “Methodological approach: Descriptive analysis” and “Topic analyzed: General” when some studies used quantitative methods (e.g., correlation analysis) across multiple fields such as 16 science, medicine, social science, and humanities fields (Kousha & Thelwall, 2017).
[Authors] “Despite the large number of studies on Wikipedia, most of them are rather ad hoc works, lacking a broader conceptual framework on which to build research around Wikipedia. The lack of such conceptual framework has hindered the development of broader research perspectives, particularly regarding the relationships of Wikipedia with Science.”

[comment]. This issue needs to be fully clarified. I could not understand the above statement because many studies have investigated Wikipedia (some of them cited here) to assess “the relationships of Wikipedia with Science” either through cited references in Wikipedia, citations to Wikipedia articles, Wikipedia credibility and usage in academia and so on. Moreover, I could not understand what authors meant by “lack of such conceptual framework has hindered the development of broader research perspectives.”
I think these statements should be explained fully in a continuous argument and not just go on to help readers understand the main discussions or arguments in the relevant contexts.
other research relevant to the topic
Nielsen, F. Å., Mietchen, D., & Willighagen, E. (2017, May). Scholia, scientometrics and wikidata. In European Semantic Web Conference (pp. 237-259). Springer, Cham.
Lewoniewski, W., Węcel, K., & Abramowicz, W. (2017, October). Analysis of references across Wikipedia languages. In International Conference on Information and Software Technologies (pp. 561-573). Springer, Cham.

it is helpful to know some examples of using Wikipedia knowledge graph for different purposes (e.g., in biomedical)
Too many graphs and tables may distract readers.
Page 8, lines, 173-177. I don’t think it is necessary to mention this here “Many of the results presented here see also
Zhang, H., Ren, Y., & Kraut, R. E. (2020, January). Mining and Predicting Temporal Patterns in the Quality Evolution of Wikipedia Articles. In HICSS (pp. 1-10).

Page 8, lines, 179-18. Makes dull reading
the section 2 is super-long, almost six pages of theoretical discussion about similarity between Wikipedia and academic articles. Need to be shortened to make the text less tedious for readers and to move on to the research questions. Theoretical text could be published separately and mixing it with the quantitative results may distract users for a journal article.
Some tables (e.g., Table 6) can be moved to the appendix section (e.g., online), too big and in some cases they have less connection to the main text.
Could Wikipedia articles domains have an impact on quantitative results, a case study in one or two domains could be interesting for the discussion.

Thank you for addressing comments and improving the readability of the manuscript.