Palak Gupta👋
Turning data into insights with my Strategic Data Analysis
Portfolio Project 3:
AirBnB Data Aalysis
Overview
The Airbnb Data Analysis project focused on exploring and uncovering insights from Airbnb listings to understand pricing patterns, customer behavior, and neighborhood trends. The project aimed to assist hosts, travelers, and stakeholders in making data-driven decisions through interactive visualizations and predictive insights.
Research: Initial research involved understanding how Airbnb operates across different cities, especially the impact of location, amenities, and reviews on listing prices and occupancy rates. The project used open-source datasets (e.g., from Inside Airbnb) with real booking data, reviews, and listing metadata.
Information Architecture: The dataset was structured around key dimensions: location (latitude, longitude, neighborhood), price, availability, reviews, and host attributes. Data cleaning and transformation ensured missing values were handled, prices were normalized, and geospatial fields were converted for mapping.
Wireframing and Prototyping: Prototypes of dashboards were designed using tools like Power BI and Tableau, allowing users to filter by city, price range, and review scores. Wireframes also included predictive models for pricing recommendations.
Challenges
Data Inconsistencies:
- Challenge: Inconsistent formats for price, missing values in reviews and amenities columns.
- Solution:Used pandas for cleaning—handled nulls, removed outliers, and standardized the price format using regular expressions.
Geospatial Visualization:
- Challenge:Visualizing listings on a map with interactive elements like price and review density.
- Solution:Used Folium and Plotly for interactive heatmaps and clustering of listings by location and popularity.
Pricing Prediction:
- Challenge:Building a reliable model to predict listing price based on features like location, room type, number of reviews, and host status.
- Solution: Applied regression models (Linear Regression, XGBoost) and evaluated them using RMSE. Feature importance was extracted and visualized.
Exploratory Depth:
- Challenge:Deriving meaningful insights beyond basic summary statistics.
- Solution:Performed correlation analysis, temporal trend evaluation (e.g., price by month), and sentiment analysis on reviews to uncover deeper patterns.
Results/Conclusion:
The analysis revealed key factors that influence Airbnb listing prices and popularity, such as proximity to tourist landmarks, host responsiveness, and review quality. Predictive pricing models achieved strong performance, helping suggest optimal pricing for new listings. Visual dashboards allowed users to interact with the data intuitively. The project not only provided business value for hosts but also demonstrated advanced data cleaning, modeling, and storytelling skills. Future work may involve integrating weather and event data to enhance booking trend predictions.
