The Evolution of the Concept of Semantic Web in the Context of Wikipedia: An Exploratory Approach to Study the Collective Conceptualization in a Digital Collaborative Environment
Xiaotian Chen
Round 1
Reviewer 1 Report
The article introduces a form of discourse analysis that was conducted on one page in the English version of Wikipedia ("Semantic Web") in order to understand changes to the definition of the term over time. In this sense, the paper is a nice example of a digital humanities project in which the cultural evolution is tracked by means of a digital artefact.
However, I think there are at the moment to many issues in all parts of the paper that it could be justified to be published. I list those issues below. I think these need to be corrected before the paper could be considered again.
(1) Improve the fit with the Special Issue.
At the moment, I think the paper is borderline as to how it relates to teh Spcial Issue as I understand it. The Special Issue asks to explore "transfigurations of well-known practices" in scholarly publishing. The paper could make a better stance as to how the analysis of Wikipedia allows a deeper analysis of how scholarly practices are transforming. Maybe what the authors are doing is related to the "science popularisation" part in scientific publishing, as Wikipedia is a source of popular science communication.
(2) Better draw out the contribution and relation to prior work
The Introduction contains a rather shallow listing of some prior work (e.g Cress model collaborative knowledge construction) without sufficient analysis of what these papers have found, and how the current draws on these results and what it will add. There is a lot of "string citing" which is not considered good practice. Maybe it would be recommednable to have teh argument about the contribution in the introduction, and then to include a second subsection called "Background" or "State of the Art" and include there the more detailed analysis of prior work.
(3) Clarify the theoretical background.
Several conceptual starting points are now mentioned in the paper, such as the work of Cress (building on Luhmann), Scardemelia & Bereiter, Hjorland and Dahlberg. I am not sure the integration of these diverse perspectives is clear in the present version. For example, what is the role of the social/collelective kowledge and where is it "situated", and what is the role of "individual knowledge" (which is very important in Cress work, but not really focused on in the present paper)
(4) Clarify the Method
It is not clear how the "pattern-matching time-series" was conducted. This needs a lot more detail to be understandable.
(5) Clarify and better justify the conclusions drawn
I am not sure that conclusions like "the search for adaptation to non-specialist readers by Wikipedia editors marks a significant difference between the two scopes (Wikipedia and Berners-Lee's publications?)" can be sustained by the data that was collected. I don't see from the analysis how the relation between the two scopes was done and how the conclusion were drawn. I also don't understand how the authors arrive at the two types of support they call "continuum" and "discrete units" and what these actually mean. As this seems to be a central conclusion, this needs to be clarified.
Author Response
(1) Improve the fit with the Special Issue.
At the moment, I think the paper is borderline as to how it relates to teh Spcial Issue as I understand it. The Special Issue asks to explore "transfigurations of well-known practices" in scholarly publishing. The paper could make a better stance as to how the analysis of Wikipedia allows a deeper analysis of how scholarly practices are transforming. Maybe what the authors are doing is related to the "science popularisation" part in scientific publishing, as Wikipedia is a source of popular science communication.
R) We agree that the subject addressed may not be the most central to academic publication, but we believe that it is relevant given the transformations of scholarly communication models and practices. We hope that the changes made in the Introduction and the approach added in the Background section (now created) reflect this perspective.
[we attach a new version of the text for you to better judge the changes that we talk about]
(2) Better draw out the contribution and relation to prior work
The Introduction contains a rather shallow listing of some prior work (e.g Cress model collaborative knowledge construction) without sufficient analysis of what these papers have found, and how the current draws on these results and what it will add. There is a lot of "string citing" which is not considered good practice. Maybe it would be recommednable to have teh argument about the contribution in the introduction, and then to include a second subsection called "Background" or "State of the Art" and include there the more detailed analysis of prior work.
R) Following the suggestion of the reviewer, a brief presentation of the paper was added to the Introduction and the section Background was created.
The list of works presented is not intended to be exhaustive, only exemplifying studies that, from different perspectives, use Wikipedia as a field of research. In addition, the focus of these works, despite some contact points, differs from that of this paper, so a more detailed analysis of these works is outside the scope of this study.
(3) Clarify the theoretical background.
Several conceptual starting points are now mentioned in the paper, such as the work of Cress (building on Luhmann), Scardemelia & Bereiter, Hjorland and Dahlberg. I am not sure the integration of these diverse perspectives is clear in the present version. For example, what is the role of the social/collelective kowledge and where is it "situated", and what is the role of "individual knowledge" (which is very important in Cress work, but not really focused on in the present paper).
R) As mentioned in the previous point, the specific aspects involved in "knowledge building" (especially "individual knowledge") go beyond the scope of the present study. Thus, to avoid too many conceptual starting points, as the reviewer pointed out, it was chosen not to refer to Cress and Kimmerle's socio-cognitive model of knowledge creation in the Background section. As for Scardemelia & Bereiter, it is their definition for the concept "collective knowledge" that is used, it is not the concept itself that is explored. Finally, Hjorland and Dahlberg are authors of reference in the area of the Knowledge Organization whose works related to the Concept Theory provide a theoretical-practical framework relevant to the present research.
(4) Clarify the Method
It is not clear how the "pattern-matching time-series" was conducted. This needs a lot more detail to be understandable.
R) The term "pattern-matching time-series" has been changed to "time-series analysis" since the former can lead readers to a relationship with statistical methods of time-series analysis. The analysis was a content analysis, looking for patterns, using chronological sequences that, according to Robert Yin (2014, p.154), fit into this analytic technique.
[Yin, R. K. (2014). Case study research: Design and methods (5th ed.). London: SAGE Publications.]
(5) Clarify and better justify the conclusions drawn
I am not sure that conclusions like "the search for adaptation to non-specialist readers by Wikipedia editors marks a significant difference between the two scopes (Wikipedia and Berners-Lee's publications?)" can be sustained by the data that was collected. I don't see from the analysis how the relation between the two scopes was done and how the conclusion were drawn. I also don't understand how the authors arrive at the two types of support they call "continuum" and "discrete units" and what these actually mean. As this seems to be a central conclusion, this needs to be clarified.
R) Berners-Lee papers were considered as "discrete units" as they are closed to changes at the time of publication. On the contrary, the Wikipedia article is considered a "continuum" because it is continuously open, since all contributions can be reversed at any time. A summary version of these considerations has been added to the text.
In fact, in the confrontation illustrated in figure 2 and subsequent discussion, there were no examples that support the conclusion referred to, so the following excerpt was added in the Discussion section:
Explanations regarding the need to adapt the vocabulary to the non-specialist user by the editors can be found in both the descriptions of the changes (available in the article history) and in the discussion page. As an example, for the first case, "skewed the defn to an outsider's (web user's) point of view" (Vanished user kijsdion3i4jf, Feb. 23, 2008); "Query users by better explaining 'to web of data that can be processed by machines'" (Quercus solaris, Oct. 13, 2017) and, for the second, "WP content is intended for a 'general audience,' the wording should reflect that "(dr.ef.tymac, Feb. 20, 2007).
Author Response File: 
Author Response.docx
Reviewer 2 Report
While the analysis and its results are not suprising or counter-intuitive it presents an original research question and a study that is reasonably well framed and carried out. The study is an interesting little case study that has a potential to become a valuable point of reference for those studying the evolution of the semantic web.
Author Response
Thank you for the evaluation.
Given that some changes have been made to the original text, we have attached a new version of the text so that you can be aware of them.
Author Response File: Author Response.docx
Reviewer 3 Report
I like the manuscript. I also agree with Reviewer 1’s comments/suggestions, and noticed that the authors have made some revisions. One point raised by Reviewer 1 may not be addressed enough: “Better draw out the contribution and relation to prior work”. The authors’ response includes “The list of works presented is not intended to be exhaustive, only exemplifying.” I would recommend that the authors be more inclusive, if possible. I also share Reviewer 1’s concern whether the manuscript fits the special issue, but would like to let the editor(s) of the special issue decide.
The manuscript still need some proof reading/copyediting. Here are some examples:
Line 17: Authors use 2 formats of dates “between 7-12-2001 and 12/31/2017”. Need to be consistent. Date also appears in Lines 228 and 245.
Line 46: Something seems missing after “such as”.
Line 48: “Academy” may not need capital A. It also appears in Line 86.
Line 57: “Semantic Web” has capital S and W most times in the article but not here (also Lines 67 and 179). Please be consistent.
Line 88: the spelling of “analyse” may need to be consistent with the same word in Line 15. “analyse” also appear later in the article.
Lines 124-125: Replace semicolons with comas.
Lines 204-206: The sentence may need revision for grammar.
Line 206: Capital T in table.
Line 269: remove the extra punctuation mark period.
Lines 270-271: “others will be necessary to understand” may need revision for grammar.
Author Response
We would like to thank the reviewers and editors and hope that the changes will meet the requests made.
To make the contributions of the work more explicit we added the following (marked in blue color) to the conclusions section:
«Regarding the relationship between Wikipedia and academia, the study points to the target audience as a relevant difference factor. This feature of moderating the language used to reach a wider audience implies that Wikipedia, even without the NOR rule, should be taken as a complementary not an alternative medium for scientific dissemination.
Given the characteristics of Wikipedia, described and discussed throughout this paper, we can consider it as a place for collective bargaining of meanings, and it is therefore important to take it as an object of study for a community's understanding of any concept in particular. This position is aligned with Hjørland quote: "Concepts have been understood as socially negotiated meanings that should be identified by studying discourses rather than by studying individual users or a priori principles." [19]. In this context, this research presents an approach, for the diachronic study of these discourses using the information source and features provided by Wikipedia.»
For study's context, in the introduction section, we added more details about the previous work related to the definition of the semantic web concept:
«One previous research, focused on the statements of the World Wide Web Consortium (W3C) and its director (Berners-Lee) works, showed that “the concept of Semantic Web is ambiguous and misinterpreting, given its biasing connection with the term ‘semantics’ and the association to other terms such as Web of Data, Linked Data or even Web of Linked Data.” [20]. The study describes the terminological and conceptual metamorphosis of the definition of Semantic Web, expressed in the documents analyzed.»
In addition, we added (also in the Introduction) a reference were one can find a systematization of the research areas of works related to Wikipedia and introduce, to the background section, the Table 1 with a selection of Wikipedia quality studies.
In respect to the specific points we have made the modifications suggested and some comments in the cases considered relevant.
Line 46: Something seems missing after “such as”.
Unless there has been a formatting error, after "as", there should be the two references of the articles with "extensive literature reviews", as it is in the file that we send:
«Despite the sheer number of works on Wikipedia, it is possible to obtain a comprehensive view of their focus from extensive literature reviews such as [10,11], or of...»
Line 57: “Semantic Web” has capital S and W most times in the article but not here (also Lines 67 and 179). Please be consistent.
Following the suggestion, we put the term semantic web (with lowercase letters) whenever we refer to the concept, and Semantic Web (with capital letters) when the Wikipedia article with that name is mentioned.
Line 88: the spelling of “analyse” may need to be consistent with the same word in Line 15. “analyse” also appear later in the article.
In the last revision we change the spelling of “analyze” to “analyse” in all occurrences, so, as in the case of line 46, the spelling should always be the same throughout the text.
Line 269: Remove the extra punctuation period.
We did not find the mentioned case.
