COVID-19 and tourism: extracting public attitudes

ABSTRACT

Taking advantage of the users’ posts on Twitter, we investigate the impact of COVID-19 on tourism in the early months of the epidemic. For this purpose, more than two million tweets published in the first months of the outbreak are analyzed. A comprehensive lexicon of keywords in the field of tourism, as well as international airlines, is collected and used for extracting tourism-related tweets. Employing a new model based on the RoBERTa language, we extract the sentiments of tweets for different countries. The results show differences in users’ positive or negative views in different countries. While in some countries, such as Germany, the public view is positive, the public view is negative in other countries, such as Russia.

KEYWORDS:

• Tourism
• sentiment analysis
• NLP
• knowledge extraction
• COVID-19

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 https://www.oxfordreference.com

2 https://wordnet.princeton.edu

3 https://en.wikipedia.org/wiki/Lists_of_airlines#By_continent

4 https://gate.ac.uk/