The aim of this project is to conduct an exploratory data analysis (EDA) on the Airbnb dataset for the city of Milan, Italy. By analyzing various aspects of the dataset, including pricing, availability, and neighborhood trends (and comparing some of those with the city of Rome), we seek to gain insights into the Milan Airbnb market and understand the factors that influence pricing and demand.
The primary dataset used for the analysis is "milan.csv", which is the cleansed dataset. The original .csv file is "milan_raw.csv". The main data source is the official Airbnb database, which is constantly updated every quarter of the year. The latest update for the database used in this EDA is March 22, 2024. The cleansed dataset, SQL and scripts for data processing and analysis are available in this repository for further exploration and utilization. For the comparison at Question 2), the dataset "rome.csv" was recovered and cleansed in the same way of the previous one.
During the cleaning and preparation phase, following tasks were performed:
- Creation of a Staging table to preserve the original data
- Identification and removal of duplicates and outliers
- Standardization of Data (formatting)
- Handling missing and null values to avoid imprecise analysis
- Excel - Initial data cleansing, addressing formatting inconsistencies and improvment of data quality
- MySQL Workbench - Data Cleaning
- MySQL Workbench - Data Analysis
- Tableau - Data Visualization for Dashboard
- R - Correlation and quick visualization
- What is the most common room type?
See the findings
The analysis reveals that the most common room type in the majority of neighborhoods is "Entire home/apt." This room type consistently appears as the dominant listing across various areas. By evaluating the count of listings per room type within each neighborhood, we observe that it has a significantly higher number of listings compared to other room types such as "Private room" or "Shared room." Also, being an entire home, its average price is higher than others because it can hosts more guests.| Room Type | Count | Average Price |
|---|---|---|
| Entire home/apt | 16485 | 177 |
| Private room | 2531 | 113 |
| Shared room | 151 | 99 |
| Hotel room | 21 | 112 |
| ... | ... | ... |
- What are the cheapest and most expensive neighbourhood in Milan? Is there a significative difference between the prices of the two most populated cities in Italy? Compare Milan and Rome.
See the findings
After finding the three most expensive and cheapest neighbourhood for each city, as we can see, there's a very high discrepancy between the two, with Milan having the highest average prices, twice as big compared to Rome, making it so most expensive to visit.
Things change for the cheapest, where the average price is higher in Rome than Milan (even though the difference from first to last is not that wide as before).
Also the number of listing is very different: Milan has very few listing for the neighbourhood, both in the cheapest and most expensive. In fact, Rome is 1st in ranking for number of AirBnB italy, so that is not surprising (number of neighbourhood groups in Rome is also lower and restricted in fewer zone of areas, so, for this cause, listing in each of those is higher).
--Milan--
| Neighbourhood | Average price | Listing count |
|---|---|---|
| Parco bosco | 63 | 8 |
| Bruzzano | 69 | 41 |
| Adriano | 70 | 69 |
| Ex om - morivione | 507 | 115 |
| Umbria - molise | 412 | 321 |
| Cantalupa | 373 | 4 |
--Rome--
| Neighbourhood | Average Price | Listing Count |
|---|---|---|
| V Prenestino/Cento | 101 | 1319 |
| VI Roma DT | 110 | 268 |
| XIV Monte Mario | 120 | 654 |
| I Centro Storico | 225 | 15861 |
| X Ostia/Acilia | 217 | 818 |
| IV Tiburtina | 167 | 560 |
- What is the busiest month of the year to visit milan?
See the findings
For this analysis, we consider the amount of reviews by month as a value that indicates the number of people who visited the city that month. We don't consider 2024 since it is still running. As we can see, the very high pitch of bookings is during september and october, but considering that reviews are published during the end of the staying, we can deduce that august and september are the busiest months of the year.
- Who are the five host with the highest number of listing?
See the findings
For this calculation, we need to consider also the host_id, otherwise hosts with the same name would be put together. The host with highest count of listing, as we can see, is "Italianway", which he appears for that reason in other several calculation.
| ID | Name | Listing Count |
|---|---|---|
| 27693585 | Italianway | 259 |
| 9025189 | Easylife House | 158 |
| 4417813 | The Best Rent | 129 |
| 16366171 | Cleanbnb House | 107 |
| 14743308 | Casa Da Suite | 85 |
From this analysis, the data available allow us to gather some more interest finding, summarized in the following dashboard, which the interactive version can be found on Tableau Public.
Given that Milan has over 80 neighborhoods, to avoid a confusing and hard-to-read displaying, certain analyses consider only the top five neighbourhoods by number of bookings.
View the Dashboard
The Buenos Aires - Venezia neighborhood has the highest number of listings, with 1,475 in total. Andrea is the top host, owning 41 of these listings. The average price for an entire home in this neighborhood is 192 euros per night (25 euros more than the average price of a staying in Milan, which is 167 euro).