Data on Sentiment Analysis of determinant factors for Airbnb accommodation by Continents

Published: 10 September 2019| Version 1 | DOI: 10.17632/7pdyztxr73.1


Airbnb is a peer-to-peer accommodation website in the sharing economy. Sentiment analysis is to identify expressions in a text to determine polarity. Intrigued by how natural language influences consumer’s choice, this study aims to examine how consumer’s contexts differ across different geographical area. A total of 8740 tweets were collected in a period of 1 month (March to April 2019) from Twitter Archiver and are classified under 6 continents; Africa, Asia, Europe, North America, South America and Oceania. Tweets were analyzed and assisted with the corpus analysis tool, WMatrix. Our main goal is to test the 8 motivations and also address the gap of existing literatures, who tend to ignore the importance of culture in choosing accommodation. Findings confirm that Airbnb users tend to consider motivations like price, social interaction, location, opinion, materialism, photo, demographic and communication while choosing Airbnb accommodation. Surprisingly, Animals (dogs, cats) and cultural differences were also identified as a strong motivation for certain users. Yet, negative sentiment from tweets was mostly caused by noise. Our results give an insight to Airbnb hosts in order to analyze their current market situation and improve their revenue. The paper also shows imperative motivations that hosts might need to consider in a guideline prototype.



Sunway University


Natural Language Processing, Continent, Twitter, Sharing Economy, Sentiment Analysis