The goal of this project is to help Airbnb hosts understand which one of six category scores contributes the most to the overall rating using linear regression.
Some Airbnb hosts complained in the Airbnb community that they received six perfect category scores but not a perfect overall rating for their listings. Overall ratings is a separate score given by guests.
To get started, datasets are available from http://insideairbnb.com/get-the-data.html. Required Python libraries are mentioned in the "requirements.txt" file.
In the combined data of San Francisco, Seattle, and New York City, cleanliness is the most important, followed by value and accuracy.
Check-in is not a big factor to the overall rating. It's possible that the duration of check-in is usually quick (less than 5 mins) and hence, it has minimal effect on guests' overall staying experience. Moreover, location is the least important when it comes to overall rating. It could be most listings' location is more or less the same as guests expected. As a result, it isn't a determining category.