Project 3

Airbnb Data Analysis project

Airbnb Data Analysis project

The Problem / Opportunity:

This project focuses on performing Data Cleaning, Preprocessing, and Market Analysis on raw Airbnb datasets (Listings, Calendar, Reviews) for Athens, Greece, using Python (Pandas & NumPy). The primary objective was to build an end-to-end data processing pipeline to clean raw strings and currency metrics, handle missing values and anomalies, and derive strategic business metrics regarding market profitability, host volume, and pricing efficiency.

Usage Instructions:

  • Clone or download the repository to your local machine.
  • Ensure Python 3.x and the required libraries (pandas, numpy) are installed.
  • Open the Airbnb_Data_Analysis.ipynb file using Jupyter Notebook or Google Colab.
  • Run all cells sequentially to execute the data cleaning pipeline and output the structured metrics and DataFrames.

Data Source:

  • The datasets (Listings.csv, Calendar.csv, Reviews.csv) were provided as part of the "Python for Data Professionals" course by Data Tutor (datatutor.gr).
  • The datasets contain detailed information about property listings, daily calendar booking status/prices, and guest reviews for the Athens area.

Corrections,Imputation and Calculations:

  • Data Pipeline Robustness: Applied data formatting (removing $, %, ,), implemented Mode Imputation for response metrics, and neutralized extreme booking placeholders (999 nights capped at 365 using .clip()).
  • Most Profitable Areas: Isolated actual bookings (available == 'f') to identify top-earning neighborhoods in Athens (e.g., Σταθμός Λαρίσης, Πεδίο Άρεως, Θησείο) based on average listing earnings.
  • Host Review Volume: Aggregated review counts across properties to highlight the most dominant hosts and property managers driving market engagement.
  • Pricing Strategy Efficiency: Developed a custom metric (Revenue per Available Day) using np.where() to prevent division-by-zero errors, revealing listings with optimal daily yields during active windows.

Business Recommendations:

  • Strategic Investment: Property investors should focus on high-yield areas like Stathmos Larisis and Thissio, which show higher average listing revenues compared to saturated central zones.
  • Pricing Optimization: Hosts should monitor Revenue per Available Day rather than raw daily rates to balance occupancy and maximize total revenue during active listing periods.

The Outcome:

Created a complete Python data processing pipeline and exploratory data analysis notebook that transforms raw Airbnb data into clean, structured DataFrames and actionable strategic insights for short-term rental management.