This project focuses on Exploratory Data Analysis (EDA) and Feature Engineering (FE) on Google Play Store application data using Python. The objective is to analyze app characteristics such as ratings, reviews, installs, pricing, and categories to understand factors that influence app popularity and performance.
The goals of this project are to:
- Analyze Google Play Store app data to identify popularity and performance drivers
- Perform EDA to understand rating, install, and pricing patterns
- Clean and standardize messy real-world app data
- Apply feature engineering to prepare data for advanced analytics or modeling
- Generate insights useful for app developers and product teams
- Google_Playstore_analysis.ipynb – Python notebook containing EDA and feature engineering
- playstore.csv – Cleaned dataset used for analysis
The dataset contains information about mobile applications listed on the Google Play Store, including:
- App Name
- Category
- Rating
- Reviews
- Size
- Installs
- Type (Free / Paid)
- Price
- Content Rating
- Genres
The data reflects common real-world challenges such as missing values, inconsistent formats, and mixed data types.
The project was executed using the following structured approach:
Reviewed dataset structure, handled missing ratings, removed duplicates, and corrected inconsistent formats in installs, price, and size columns.
Analyzed distributions and relationships across:
- App ratings and reviews
- Install counts
- Free vs paid apps
- Category-wise app performance
Performed transformations such as:
- Converting installs and price into numeric values
- Encoding categorical variables
- Creating derived features for analysis readiness
Ensured all features were structured, clean, and suitable for further modeling or dashboarding.
- Majority of apps on the Play Store are free
- Apps with ratings between 4.0 and 4.5 dominate the platform
- High install counts do not always guarantee higher ratings
- Certain categories consistently attract more users and installs
- Data cleaning significantly improves analysis reliability
- Real-world EDA using Python
- Handling messy, high-cardinality categorical data
- Feature engineering on numeric and categorical attributes
- Business-focused interpretation of app performance
- Model-ready dataset preparation
- Strong foundation for recommendation or prediction systems
- Free apps dominate the market in terms of installs
- Paid apps do not necessarily achieve higher ratings
- App category plays a major role in user engagement
- Ratings alone are insufficient to judge app success without install context
- Perform sentiment analysis using app reviews
- Build predictive models for app ratings or installs
- Develop a recommendation system based on app features
- Visualize insights using Power BI or Tableau
- Track app performance trends over time
Tejeshwar R
Data Analyst | Python | SQL | Power BI | EXcel | Statistics