Investigating the Impacting Factors on the Public’s Attitudes towards Autonomous Vehicles Using Sentiment Analysis from Social Media Data
Round 1
Reviewer 1 Report
1. The research primarily focuses on the description and fitting analysis of the data; nevertheless, there is minimal explanation of the impact's causality.
2. The study's objectives need to be clarified more clearly, and practice's importance should be explained in more detail.
3. This study's independent variables are Avs-related keywords, and the dependent variable is the sentiment score. Please explain further and discuss the logical relationship between the independent and dependent variables.
4. No clear research conclusions have been summarized, which should be further arranged and summarized.
5. Figure 2 seems not meaningful to the research content.
6. Figure 3 get the importance of each variable through random forest regression. What is the difference comparing with ANOVA? What is the advantage?
7. Lines 161-167: Why are topics and variables divided into seven categories? Is there any theoretical basis?
8. Line 335-336, my understanding is that you ran the linear mixed model twice, the first one was to judge the valid variables, and the second one was to estimate the coefficients (Table 3). If so, kindly provide a detailed description of the process.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
I believe this is an interesting and very relevant study, especially for Autonomous Vehicles (AVs) manufacturers, relevant policy makers, etc. In this study, the authors aimed to investigate the impacting factors on the public’s attitudes towards AVs from social media data. The results showed that the attitudes of the public toward AVs were slightly optimistic. Factors like “drunk”, “blind spot” and “mobility” had the largest impacts on public attitudes. In addition, people were more likely to express positive feelings when talking about words such as “lidar” and “Tesla” related to high technologies. Conversely, factors such as “COVID-19”, “pedestrian”, “sleepy”, and “highway” were found to have significantly negative effects on the attitudes of the public.
While this article is worthy of publication, I do have the following concerns：
1. While this article is novel, it is not a groundbreaking study on public attitudes toward AVs.
For example, a related study is "The Acceptance of Independent Autonomous Vehicles and Cooperative Vehicle-Highway Autonomous Vehicles [J]. Information, 2021, 12(9): 346."
Therefore, the author's reference to "To fill this research gap" in lines 110 and 111 is inappropriate.
2. Various methods are used in the paper, such as Sentiment analysis, Random forest and Linear mixed model. The interest of potential readers can be more aroused if the logical relationship between different methods and the computational flow of each method can be shown with graphics.
3. The authors should strive to do a closer analysis and discussion of the results calculated by the model. For example, what might be the underlying reasons for the volatile change over time in public attitudes towards AVs?
4. The authors should conclude by stating the inadequacies of the study.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
1) The article presents interesting research and very well prepared. The test procedure is clear and justified. The study data are rightly chosen and sufficient. The authors of the article correctly described the study; it is important that the test procedure can be reproduced by other researchers.
2) This is an interesting paper dealing with Social media, which today plays a very important role in a number of areas of human life, it is therefore suitable for research in the field of social sentiment
3) What was criterion of selection dates Jan 2019-Nov 2020? Why did not you used e.g. whole 3 years to analyse yearly trends/changes?
4) Is there a specific list of the keywords used in snscrape? You mentioned “several AV related”, how many keywords you used
5) It is worth to show impact on the science of social media in other than social studies, I recommend for a references to another interesting research, e.g.:
a) Kapidzic S, Neuberger C, Frey F, et al (2022) How News Websites Refer to Twitter: A Content Analysis of Twitter Sources in Journalism. Journal Stud 23:1247–1268. https://doi.org/10.1080/1461670X.2022.2078400
b) Mpofu P, Asak MO, Salawu A (2022) Facebook groups as transnational counter public sphere for diasporic communities. Cogent Arts Humanit 9:. https://doi.org/10.1080/23311983.2022.2027598
c) Wengel Y, Ma L, Ma Y, et al (2022) The TikTok effect on destination development: Famous overnight, now what? J Outdoor Recreat Tour 37:100458. https://doi.org/10.1016/j.jort.2021.100458
6) Part of the text is not prepared according to the instruction for the authors, e.g. bottom part of the Table 2.
7) In my opinion you need to very clearly state (in the text) why did you used random forest method instead of any other for the calculation variables importance
8) However, the last but most important note: the authors only used data from Twitter using the SNScrape library. What was the criterion for choosing only Twitter and not other, more popular social media? Looking at the list of social media with the most monthly logins Twitter is only 17th (https://en.wikipedia.org/wiki/List_of_social_platforms_with_at_least_100_million_active_users#cite_note-:1-1). Please refer to this very clearly in the revised text. In my opinion, it would be more expedient to pore over 3-5 social media and compare the results. E.g. Reddit users may use completely different variables/phrases.
Text might be accepted after improvements according to the above remarks.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
Overall, the manuscript presents a good research objective, with results and methods adequate to said objectives. Perhaps the conclusion should be completed.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
The authors corrected the text according to the suggestions of all reviewers, in its current form it is suitable for publication.
Author Response
We would like to thank the anonymous reviewer for your recommendation for publication and thoughtful suggestions that have helped to improve this paper substantially.
