Trust-Aware Evidence Reasoning and Spatiotemporal Feature Aggregation for Explainable Fake News Detection
Round 1
Reviewer 1 Report
The submitted manuscript is devoted to the topical topic of detecting fake news. The work considers an innovative and popular approach based on trust.
1. The Introduction section of the manuscript is clear, and the scientific contributions at the end of the introduction are thoroughly presented.
2. The literature review is comprehensive and shows the proper place of the proposed approach. The sources are primarily new.
3. The theoretical section is performed thoroughly and transparently.
4. The Experiments section contains discussion elements - it should be called Experiments and Discussions instead, with information about the limitations of the proposed approach.
5. The conclusion section should be expanded using 1) numerical results obtained in work and 2) limitations of the proposed approach.
In sum, the submitted manuscript can be accepted after minor revisions based on the reviewer’s comments.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 2 Report
Comments and Suggestions for Authors
Thank you for the opportunity of reviewing this interesting article. This study develop a trust-aware evidence reasoning and spatiotemporal feature aggregation model for more interpretable and accurate fake news detection. Autors design a trust-aware evidence reasoning module to calculate the credibility of posts based on a random walk model to discover high-quality evidence. Next, from the perspective of spatiotemporal structure, Autors design an evidence-representation module to capture the semantic interactions granularly and enhance the reliable representation of evidence. Finally, a two-layer capsule network is designed to aggregate the implicit bias in evidence while capturing the false portions of source information in a transparent and interpretable manner. This approach seems to be extremely relevant and promising.
Comments and Suggestions for Authors:
· The authors indicate: «We propose an interpretable fake news detection method called TRSA based on trustaware evidence reasoning and spatiotemporal feature aggregation». Meanwhile, the base on which the testing was carried out is not described, nor are specific examples indicated.
The absence of a description of the base significantly reduces the significance of the work.
· Please describe in more detail the tools you used in your research.
· The authors indicate: «To unfold user attention distribution differences between fake and true news content, 600 we randomly select three fake (0–2) and three true (3–5) news stories, and plot their 601 token weights distribution based on the attention of the interactions between the evidence and claims». Please explain how true or false was determined in each case.
· Please adjust the structure of the article according to the requirements.
· The process of discussing the results can be extended by applying the results and extrapolating them to other similar studies.
· Please describe in detail how your study fits for aims and scope of Applied Sciences.
· For theoretical framework and bibliography additional current references should be included to new research 2022-2023.
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
The paper is a very good and addresses a very important topic. The problem setting, hypothesis and methodology part is very good, as well as, the paper has an in-depth literature review. I have one three minor suggestions for the authors to consider:
1. I would like to see some policy implications for this study. Since this is such an important and valid topic, it has a strong policy significance, and specific actions which can be taken by the government, especially elaborating on early detection of fake news, which helps to manage any emergency situation well.
2. The other point is, it will be good to see some discussion on the role of new and emerging technologies, especially AI and big data analysis on reducing the impacts of fake news
3. The third point is to enhance trustable sources more to negate the impact of fake news.
Apart from the above three points, I found the paper is extremely good, and can be published with minor corrections.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 4 Report
The theme is interesting, but some points need to be clarified better.
In the introduction, make it clear what is new in the article
A related work section is imperative, discussing and comparing the novelties and problems solved in this work with the current literature in relation to what already exists.
This work https://doi.org/10.3390/s22218292 may be useful to authors has the same theme and were recently published by the same publisher. https://doi.org/10.3390/s20216030
The definition of variables used in the tests is not clear why and the parameters used;
I strongly suggest describing the scientific contributions in the conclusions and defining clear and precise points readers can use in future works.
I suggest including a list of mathematical symbols and abbreviations
Review references, several are not completely missing DOI or ISSN.
Adapt the text and format according to the journal's template.
I hope I have contributed to the improvement of the article.
Author Response
Please see the attachment
Author Response File: Author Response.docx