Analyzing Public Opinions Regarding Virtual Tourism in the Context of COVID-19: Unidirectional vs. 360-Degree Videos

Round 1

Reviewer 1 Report (Previous Reviewer 2)

You wrote, “It mimics, duplicates, and even outperforms the allure of virtual exploration via the Internet or virtual technologies by creating an online environment that allows people to enjoy immersive travel while at home” Ambiguous sentence, please correct

 

You wrote “As mentioned in section 6.3, we use the SentiStrength-SE and SenticNet7 to analysis the sentiment polarity in the next step.”, apparently you never mention SenticNet7in section 6.3. Please explain in detail what is SenticNet7 and how to use it.

 

It is better if you add at least an illustration example of what the results of the polarity classification are and what the results of the sentiment analysis stage are. what's the difference between them? Why is there a need for sentiment analysis, if the polarity is already clear? Therefore, a step-by-step example based on dummy data from beginning to end will be very good for the understanding of the reader.

 

Explain why you chose CNN. Why not choose another method like YOLO, etc? Give a strong reason along with valid academic references, so don't sound like cherry-picking.

 

Explain what you update to the CNN that you use. If you only use it and do a few folds, I don't think there will be any contribution. Explain in detail what your contribution to CNN is.

 

Your discussion is just like delivering results. The Discussion chapter should contain an in-depth interpretation of the results of your research. Compare with the results of previous studies. Give reasonable insight. Provide some future direction. The discussion section is the most important part of a scientific paper because it shows the depth and breadth of the researcher in understanding the research topic. At this point, it appears that your paper is very under-contributed. And don't forget to adapt your discussion to the fix I asked for, i.e. that you don't just compare the 5 fold. But you have to compare it with other similar previous studies and have to update the recent state-of-the-art method, giving a new twist to the current method.

Author Response

Dear Reviewer 1,

First of all, we thank you for your review and valuable comments, and we thank you so much for your reminder. We carefully considered the notes from you and have already updated our article.

In addition, we also invited Professor Petr Silhavy ([email protected]) and Professor  Radek Silhavy ([email protected]) to join our article as co-authors, and Professor Petr Silhavy is a new correspondence.  We have included the "authorship change form" in this zip file. This zip contains four files, including the authorship change form; ResponseReviewer; Tex file, and article-with track_changes", and the zip file has already been submitted to Assitant Editor.

Thank you so much!

Best Regards,

Huynh Thai Hoc 

Author Response File: Author Response.pdf

Reviewer 2 Report (Previous Reviewer 1)

The revised version of the paper covers the main weaknesses mentioned in the previous revision. Authors have performed the experiments with two alternative sentiment analysis tools, demonstrating the superiority of one of them. Authors have also take into consideration CNN models which also shown a good performance compared to the previous version.

Overall, the paper now shows now a much strong conclusions, and it is ready for publication after some minor revision for typos.

Page 10. Table 5 caption includes the command cyan.

Author Response

Dear Reviewer 2,

First of all, we thank you for your review and valuable comments, and we thank you so much for your reminder. We carefully considered the notes from you and have already updated our article.

In addition, we also invited Professor Petr Silhavy ([email protected]) and Professor  Radek Silhavy ([email protected]) to join our article as co-authors, and Professor Petr Silhavy is a new correspondence.  We have included the "authorship change form" in this zip file. This zip contains four files, including the authorship change form; ResponseReviewer; Tex file, and article-with track_changes", and the zip file has already been submitted to Assitant Editor.

Thank you so much!

Best Regards,

Huynh Thai Hoc 

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report (Previous Reviewer 2)

I appreciate your effort to do the revision. However, in the discussion section, please compare your research results with other similar previous studies. 

Author Response

Dear Reviewer,

Thank you for your review and valuable comments. We have already incorporated all your recommendations into the discussion section. We appreciate your attitude and valuable contribution.

Thank you so much!

Thanks & Best Regards,

Author Response File: Author Response.pdf

Round 3

Reviewer 1 Report (Previous Reviewer 2)

Please check grammatical errors

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

This paper presents a study about the sentiments expressed by potential tourists during tge COVID-19 pandemic movility restrictions. This stady aims at determining the impact of two video modalities: 360º vs. unidirectional. The main conclusion of the study is that most opinions are neutral and therefore there is low impact in the 360º presentations.

The paper has several serious drawbacks that make this paper not recommedable for publication:

1) Authors do not take into account modern techniques for sentimen analysis. Apart from current transformed-based pre-trained models that can be directly applied to the texts, new dictionary methods like SenticNet7 has shown to largely overcome methids like SentiStrength. This is a relevant issue since the majority results in "neutral" class can imply that the method is unable to identify the right polarity of the comments due to dictionary limitations. 

2) It is unclear why authors introduce automatic classifiers, when the study is driven by the dictionary-based method of SenticStrength. Moreover, the there is no description of these classifiers: are they binary or multi-class? have the authors performed a k-fold evaluation? Which is rate of the split of train-test? Which is the kernel of the SVM?

3) How authors interpret the results of the classifiers? The only possible conclusion from the best results is that most of the SentiStrength-based labels are easily separable. Why authors do not use the final predicions of these classifiers to do a similar comparison to Table 3?

Other minor comment is that Figure 4 is not necessary, as the other classification methods have not been graphically depicted in the paper.

Reviewer 2 Report

You wrote, “It mimics, duplicates, and even outperforms the allure of virtual exploration via the Internet or virtual technologies by creating an online environment that allows people to enjoy immersive travel while at home” Ambiguous sentence, please correct

 

You wrote, “According to several advisors, virtual tourism has developed a new tourist model with several benefits, for example, recreating initial chronological arrival of a destination and preserving intangible heritage”. Please explain who are the advisors. What is their specialty that deserves to be asked opinion?

 

You wrote, “In the prior literature, virtual travel problems were centered on non-crisis and ordinary scenarios.” Please cite which literature. And what was the result like? So every claim you make should be supported by clear academic sources

 

Place Figure 1 in the position after it is mentioned in the manuscript. So don't put it before

 

The background section is still not sharp, why should comments from travelers be analyzed?

 

Then why Youtube? Why not other social media, for example, Instagram? Provide an appropriate scientific basis for the choices you make.

 

Then it's still not clear, why should comments on the video? Why not comments on photos? Your background is still very unclear.

 

In line 42, you said, “What is the proper perception of travelers toward virtual travel in a crisis?” So, how do you ensure that the comments are really from travelers? It could be that those comments are only from ordinary youtube users and not travelers. So how do you validate your data?

 

In line 136, you said” As mentioned in the Introduction section, we studied six videos from..” however, apparently you never mentioned that in the introduction.

 

You said in Line 176-178 that “According to [33], SentiStrength measures two types of sentiment: -1 (not negative) to -5 (extremely negative) and 1 (not positive) to 5 (strongly positive)”. It can be seen in the sentence that there are only 4 classes. But why in Table 2 there are 5 classes? You must explain in detail in your manuscript.

 

Please explain in detail what Sentiment Analysis is for if previously each polarity comment has been classified in the previous stage.

 

then explained in more detail how the composition of the testing data and training data. And what is the target classification? MLP, Naïve Bayes and all those algorithms you use are for classification. So what exactly is classified by these methods, if it has previously been classified in the Polarity classification stage?

 

Explain why you chose Naive Bayes, Decision Tree, Random Forest, Support Vector Machine, Gradient Boosting, and Sequence to Multiplayer Perceptron. Those are all old methods. Why don't you choose a recent method? For example CNN, VAE, GAN, etc.

 

Moreover, just comparing the existing algorithms, the contribution is very small. Therefore I ask you to choose one of the recent methods. Explain in detail why you chose that method. Then you update the method, so that is become your contribution.

 

Your discussion is just like delivering results. The Discussion chapter should contain an in-depth interpretation of the results of your research. Compare with the results of previous studies. Give reasonable insight. Provide some future direction. The discussion section is the most important part of a scientific paper because it shows the depth and breadth of the researcher in understanding the research topic. At this point, it appears that your paper is very under-contributed because this section does not exist. And don't forget to adapt your discussion to the fix I asked for, i.e. that you don't just compare the old methods that already exist. But you have to update the recent state-of-the-art method, giving a new twist to the current method.

 

In this current form, there are still many shortcomings in your manuscript. The contributions are very low. If there are no significant changes, it is most likely that your manuscript can be rejected