This project focuses on cleaning, exploring, and visualizing the Airbnb Open Dataset.
The goal is to demonstrate a data analyst–oriented workflow, starting from raw data and ending with clear, interpretable insights supported by visualizations.
Instead of building predictive models, this project emphasizes:
- data cleaning
- exploratory data analysis (EDA)
- effective data visualization
- Source: Airbnb Open Data
- Content: Airbnb listings with pricing, location, and availability information
- Main features:
room type(Entire home/apt, Private room, Shared room)priceneighbourhood groupminimum nightsavailability 365reviews per monthinstant_bookablelat/long(geographic coordinates)
The raw dataset contains missing values, inconsistent formats, and invalid data ranges.
The following cleaning steps were applied:
- Removed duplicate records
- Standardized column names (trim whitespace, lowercase)
- Converted currency-like fields (
price,service fee) to numeric format - Handled missing values using robust statistics (median imputation)
- Filtered invalid/out-of-range values:
availability 365outside 0–365minimum nightsoutside 1–365
- Dropped non-essential columns that were not useful for analysis
The focus was on clarity, reproducibility, and realistic data cleaning decisions.
After cleaning, exploratory analysis was performed to understand patterns and trends in the data.
- How is price distributed across different room types?
- Which neighbourhood groups have the highest prices?
- How does availability affect pricing?
- What is the relationship between minimum nights and price?
- Are there geographic patterns in pricing?
Price distribution shows a right-skewed pattern with most listings concentrated in the lower price range.
Entire homes/apartments command significantly higher median prices compared to private or shared rooms.
Certain neighbourhood groups show consistently higher median prices, reflecting location-based demand.
The majority of listings are entire homes/apartments, followed by private rooms.
Review counts by neighbourhood reveal which areas have the highest demand and guest activity.
Instant bookable listings tend to be priced higher, suggesting a premium for convenience.
Hosts with multiple listings employ different pricing strategies compared to single-listing hosts.
Boxplot analysis reveals outliers in review frequency, helping identify highly active listings.
- Room type and location have the most significant impact on pricing
- Instant bookable listings tend to be priced higher, suggesting a premium for convenience
- Listings with lower availability tend to be more expensive, indicating higher demand
- Central locations show both higher prices and higher demand (more reviews)
- Multi-host strategy: Users with multiple listings employ more competitive pricing strategies
- Price distribution is right-skewed with most listings in the affordable range
- Entire homes/apartments dominate the market and command premium prices
- Python
- Pandas
- NumPy
- Matplotlib
- Seaborn
- Jupyter Notebook
- Cleaning real-world datasets with missing and inconsistent values
- Converting currency strings to numeric data types
- Handling outliers and invalid data ranges
- Structuring an EDA-focused data analysis project
- Communicating insights through clear visualizations
- Working with geographic data (latitude/longitude)
📦 airbnb-data-analysis
┣ 📁 images
┃ ┣ demand_by_neighbourhood.png
┃ ┣ median_price_by_neighbourhood.png
┃ ┣ median_price_by_room_type.png
┃ ┣ multihost_pricing_behavior.png
┃ ┣ price_distribution_histogram.png
┃ ┣ price_vs_instant_bookable.png
┃ ┣ reviews_per_month_outlier_boxplot.png
┃ ┗ room_type_distribution.png
┣ 📄 notebook.ipynb
┣ 📄 Airbnb_Open_Data.csv
┗ 📄 README.md
- Open
notebook.ipynbin VS Code or Jupyter - Run cells from top to bottom
- All visualizations will be generated and saved to the
imagesfolder