Research on Optimization of Public Opinion Supervision Model of Social Network Platform Based on Evolutionary Game
Round 1
Reviewer 1 Report
The methodology of the study is not sufficiently disclosed. The objectives of the study are not clearly defined. The models given in the text of the article are insufficiently substantiated. The results of the experiment are not fully disclosed and have not been interpreted. The conclusions made by the authors in the article do not confirm the solution of the tasks set.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
This paper is motivated by the dynamics of online information spreading, and proposes an evolutionary model among three players: marketing account owners, users and platform managers.
While interesting, I think the paper needs a more solid justification. In the first place, I wonder why not using a conventional game theoretical model. Under the light of the results presented, the real-world implications of stability characterization of equilibria are not discussed; then I would think that a time invariant game theoretical model might suffice. In addition, replication equations need a justification beyond the mere mathematical convenience and should be linked to bounded-rationality concepts of strategy updating.
My second concern is the relevance of results in their implications on practical decision-making and even on further empirical studies. The authors should make an effort to translate their results to real-world cases, and convince the readers about the contribution of the model in studying this type of cases. It would be advisable in the authors draw a parallel of the model's results with a real-world case.
My third concern is the apparently trivial outcomes of the model. Of course, incrementing publishing / penalty costs will negatively affect the frequency of publishing. I think the model might tell much more than that.
Last, two minor comments: (i) the hypotheses presented on page 4 are not really hypotheses--in the empirical sense--but model descriptions. Please remove the word and rewrite; (ii) the paper needs an English style revision.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
Dear authors
The paper addresses an interesting and significant topic. However, it was difficult for me to follow the paper. You have to improve the structure of the paper.
Abstract: Introduce the topic. State clearly the purpose. Summarize the method, results, conclusions, and implications.
Introduction: Indicate the field of the work and why it is important. Provide a brief and relevant review of the literature. Indicate the research gap. Outline the purpose, state the method, and at the end give the structure of your paper.
The literature review is totally missing. You refer to the research background in the introduction. Please make a separate section and subsections and provide the relative literature.
Section 2.2. Assumptions, I think should be renamed to hypotheses. Please state clearly the hypothesis. I really don’t understand the hypothesis of the paper.
In the conclusions section please answer the hypothesis one by one clearly.
In my opinion, the most important part of your paper must be the implications. You write implications in “The Impact Analysis of Supervisory Relevant Parameters on the ESS section”. Please summarize the implications in the conclusions section.
In the conclusions, section write also the limitations of your study.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
I thank the authors for the revised version of the paper. They have satisfactorily addressed all my concerns. Nonetheless, playing the devil's advocate, I would still require a more precise discussion of some of the results.
In particular, the authors claim that information authenticity is a needed for the marketing agent in order to execute the "publish" strategy. Figure 4 ilustrates that values well below 0.5 force the marketing account to not publish. This rather implies that only high authenticity would lead an action in favor of publishing. However, in contrast, other models and research show the prevalence of diffusion of information with dubious authenticity (cf. León-Medina, F. J., Tena-Sánchez, J., & Miguel, F. J. (2020). Fakers becoming believers: how opinion dynamics are shaped by preference falsification, impression management and coherence heuristics. Quality & Quantity, 54(2), 385-412; Altay, S., Berriche, M., & Acerbi, A. (2023). Misinformation on misinformation: Conceptual and methodological challenges. Social Media+ Society, 9(1), 20563051221150412). Please provide an extended explanation.
Also, I would recomment appending the agent related strategy to all figures. For instance, Figure 3 shows the trend of probabilities for all game actors, but it does not show the associated strategy these probabilities are attached to (probability of publishing? probability of participating? etc).
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
I think that you have addressed the majority of my comments and you have answered why you keep the word assumptions. Thank you. Well done.
Author Response
Dear Reviewer:
Thank you very much for your previous comments and suggestions on our manuscript. They have greatly improved our manuscript. Thanks again!
Sincerity,
Xiaoxiao Zhai
Round 3
Reviewer 2 Report
I think the authors did not properly addressed my concerns. Again, I am rephrasing them in the following lines:
1o. The authors claim that information authenticity is a needed for the marketing agent in order to execute the "publish" strategy (in equilibrium). Figure 4 illustrates that values well below 0.5 force the marketing account to not publish. This rather implies that only high authenticity would lead an action in favor of publishing; yet, this result is in contrast with other models that show the prevalence of the diffusion of information with dubious authenticity (cf. León-Medina, F. J., Tena-Sánchez, J., & Miguel, F. J. (2020). Fakers becoming believers: how opinion dynamics are shaped by preference falsification, impression management and coherence heuristics. Quality & Quantity, 54(2), 385-412; Altay, S., Berriche, M., & Acerbi, A. (2023). Misinformation on misinformation: Conceptual and methodological challenges. Social Media+ Society, 9(1), 20563051221150412).
Please compare your results with the above-mentioned works and provide an extended explanation. In particular, how does your model results compare with claims of the above references? Explain why many examples in real life do not appear to comply with the fact that authenticity is a precondition for information diffusion.
2o. Please modify the labels of all the figures for easier reading. For instance, in figure 4, the labels in the vertical axes should say "probability of publishing", "probability of participating", and "probability of positive supervision", respectively.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
