Social Media Data-Based Sentiment Analysis of Tourists’ Air Quality Perceptions
Round 1
Reviewer 1 Report
Dear authors,
This original research paper analyses very interesting and actual topic: tourists’ perceptions of air quality. In this innovative study, using the web crawler technique, authors collected 27,500 comments on the air quality of 195 of China’s 13 Class 5A tourist destinations posted by tourists on Sina Weibo from January 2011 to December 2017, which were then conducted with a content analysis using the Gooseeker, ROST CM and BosonNLP tools, and based on the analysis of the proportions of sentences with different emotional polarities with ROST EA. As well, authors measured the sentiment value of texts using the machine learning method of an artificial neural network (ANN) through a Chinese social media data-oriented Boson platform via Python programming. Authors notice, that using sentiment analysis of social media data to understand tourists' perception of air quality has been covered in a few environmental thematic studies, and a small number of studies focusing only on this environmental component have emerged.
As well, authors notice that the paper provides evidence for new data and method for air quality research in tourist destinations and provides a new tool for air quality monitoring. So, the paper is a the contribution to the innovative tourism research, it is well written and quite clearly structured. As well I agree with the authors that „in the future, the inclusion of data from longer time period or the addition of more 615 spatial subjects can help draw more reasonable and complete conclusions“.
And I would like to share with authors some doubts too: it seems important to notice that the abstract is not very clear, and maybe would be better to structure two seperate parts of Conclusions (5.1) and Discussions (5.2) in one chapter „Discusion and Conclusions“?; the used methodology is not mentioned here very clearly; maybe it would be better to make the seperate limitations chapter (discussing critical aspects of used methodology as well, it would be interesting to know deeper authors oppinion about sentiment analysis and possible complexity of approaches using „Big Data“ too).
Best regards to authors using this novel methodology!
Author Response
Thank you for your encouragement from the reviewer. Your valuable opinions have great guiding significance for our future research. The sentence “Further exploration of the theoretical basis of the semisupervised ANT method and the introduction of other machine learning methods, using more data sources, will help to more deeply describe the phenomenon and even the mechanism analysis behind the phenomenon., mining more effective data with spatiotemporal message tags, will help to achieve a more profound phenomenon description and the mechanism analysis behind the phenomenon” has been added to the abstract. Please review, thank you again!
Reviewer 2 Report
Interesting well written paper Contribution well pointed out, although the paper is dealing with topic that has already been researched and studied / literature review Extensive literature review Good research concept/ design Well presented results
Author Response
Thank you for your confirmation by the reviewer. Your encouragement is the driving force behind our continued progress. We wish you a smooth job and good luck!
Reviewer 3 Report
The study is relevant to tourism sustainability issues, encompasses a large amount of information and uses sophisticated and pertinent methods of analysis. Methodology and results are clearly presented and reasonably discussed.
I would suggest the authors to pay attention to page 13, table 4 and its respective discussion: why is it that "the lowest quarterly sentiment value was 0.768"? There are inferior values (e.g.: 0.764; 0.757). Either there is a mistake or the analysis was not clear enough.
The research, has the authors acknowledge, should be extended and would be enriched if, complementing the interesting descriptive statistics presented, more information statistical analysis were used in the future.
Author Response
Point 1: I would suggest the authors to pay attention to page 13, table 4 and its respective discussion: why is it that "the lowest quarterly sentiment value was 0.768"? There are inferior values (e.g.: 0.764; 0.757). Either there is a mistake or the analysis was not clear enough.
Response 1: The analysis was not clear enough. I am grateful to the reviewer for your observation and accurate pointing out the shortcomings of the paper. Yes, the lowest quarterly sentiment value was 0.757. The reason is that due to the influence of inertial thinking, we believe that the lowest value comes from the 4th quarter, actually from the 1st quarter. Your rigorous academic attitude is always worth learning. Thank you again for your report!
Point 2: The research, has the authors acknowledge, should be extended and would be enriched if, complementing the interesting descriptive statistics presented, more information statistical analysis were used in the future.
Response 2: We are very happy that the paper can be of interest to reviewer. Indeed, more data sources and data with more powerful space-time tags help drive further research.The data used in the thesis mainly comes from the national, regional, provincial and scenic areas. Among them, the national and regional data are reflected in the text. The attachments are affixed with English data of 31 provinces and original Chinese data of 195 scenic spots. Please give pointers! Thank you again for your insight! Please see the attachment.
Round 2
Reviewer 1 Report
Congratulations for your efforts to review the article!
Author Response
According to your valuable comments, we have tried to reduce some of the run-on sentences and the acronyms, and added more than 700 words in the conclusion and discussion section. Due to the use of the added content for 4 days, and the content took more than 10 days for an English expert for translation, the minor revision time exceeded the 5 days specified by the editorial department. Sorry again, please forgive me. I wish you a smooth job, happy and healthy!
Tao Yuguo
