Analyzing Airbnb Data

Jonathan Trajkovic
on July 16, 2015

Editor's Note: Jonathan Trajkovic is a Data Analyst working for Synaltic in Paris, France. In this #TravelMonth blog post, Jonathan explains how he built an Airbnb viz to figure out the best place to stay in Luxembourg. Check out more of Jonathan's work on his blog, Tips and Viz with Tableau!

Thanks to Jewel Loree from Tableau Public, I found a dataset about Airbnb. The timing was excellent because I had to choose an Airbnb accomodation for a training in Luxembourg a few weeks ago. When I discovered the website Inside Airbnb, I was surprised to find many CSV files concerning several cities around the world.

As a Frenchman, I chose to start my analysis with Paris, but I did download all the location files to build the visualisation below.

The visualisation is built in two parts. The first part is a "Global view", to present the data from 10 cities with a map and a radar graph. The map is available in two types: by neighbourhood and by hexbin. To draw the hexbin, I used the method described by The Last Data Bender, which uses polygons instead of custom shapes (I think this method is more flexible and beautiful).

The second part is about hosts. I chose to show how the accommodations are distributed by hosts and by city. In this part, each accommodation is located on a map and the Airbnb URL is available.

Don't hesitate to leave a message or a note about this work.


Hi, Jonathan, for each city and each part of a city, you've analyzed Airbnb quality in five dimensions: location, accuracy, cleanliness, communications and check-in. Did you just calculate the average of theses scores for each accommodation? Perhaps you can put a filter which allows us to find out accommodations with scores above 9.5, for example. Juste un conseil!~

Merci beaucoup pour ta viz!~

Add new comment