📺 Content Monetization Modeler

📌 Project Overview

The Content Monetization Modeler is a Machine Learning project designed to predict YouTube ad revenue based on video performance metrics, audience engagement, and channel statistics.

The project utilizes data preprocessing, feature engineering, regression modeling, and Streamlit deployment to provide accurate revenue predictions for content creators and digital marketers.

---

📸 Application Screenshots

🏠 Home Page

💰 Predict Revenue Page

✅ Prediction Result

📊 Visualizations Dashboard

---

🚀 Features

• Data Cleaning & Preprocessing
• Feature Engineering
• Multiple Regression Models
• Model Evaluation using:
  • R² Score
  • Mean Absolute Error (MAE)
  • Root Mean Squared Error (RMSE)
• Feature Importance Analysis
• Interactive Streamlit Web Application
• Revenue Prediction in Real Time
• Business Insights & Visualizations

---

📂 Dataset Features

Numerical Features

• Views
• Likes
• Comments
• Watch Time Minutes
• Video Length Minutes
• Subscribers
• Year
• Day

Categorical Features

• Category
• Device
• Country
• Month

Target Variable

• Ad Revenue (USD)

---

🛠 Technologies Used

Technology	Purpose
Python	Programming Language
Pandas	Data Manipulation
NumPy	Numerical Operations
Scikit-Learn	Machine Learning
Streamlit	Web Application
Matplotlib	Data Visualization
Pickle	Model Serialization
GitHub	Version Control

---

🤖 Machine Learning Models

The following regression algorithms were implemented and compared:

1. Linear Regression
2. Decision Tree Regressor
3. Random Forest Regressor
4. Gradient Boosting Regressor
5. K-Nearest Neighbors (KNN) Regressor

---

📊 Model Performance

Model	R² Score	MAE	RMSE
Linear Regression	0.9519	3.17	13.60
Gradient Boosting Regressor	0.9517	3.69	13.64
Random Forest Regressor	0.9510	3.99	13.74
Decision Tree Regressor	0.8921	5.79	20.38
KNN Regressor	0.6622	29.06	36.06

🏆 Best Model

Linear Regression

• R² Score: 0.9519
• MAE: 3.17
• RMSE: 13.60

---

📈 Key Insights

• Watch Time Minutes is the strongest predictor of ad revenue.
• Likes and comments significantly improve monetization.
• Higher audience engagement leads to higher earnings.
• Views positively impact revenue but are less influential than watch time.
• Seasonal effects were observed during certain months.
• Audience engagement is more important than views alone.

---

📊 Visualizations Included

• Feature Importance Analysis
• Regression Model Comparison
• Revenue Driver Analysis
• Revenue Growth Trend
• Model Performance Dashboard

---

📁 Project Structure

CONTENT MONETIZATION MODELER PROJECT/
│
├── Report/
│   └── Content Monetization Modeler Report.pdf
│
├── Screenshots/
│   ├── Home Page.png
│   ├── Prediction Page.png
│   ├── Prediction Result.png
│   └── Visualizations Page.png
│
├── model.py
├── monetization analysis.ipynb
├── README.md
├── requirements.txt
├── scaler.pkl
├── youtube_ad_revenue_dataset.csv
└── youtube_ad_revenue_model.pkl

---

▶️ How to Run the Application

Clone Repository

git clone https://github.com/yourusername/content-monetization-modeler.git
cd content-monetization-modeler

Install Dependencies

pip install -r requirements.txt

Run Streamlit App

streamlit run model.py

or

python -m streamlit run model.py

---

📦 Requirements

streamlit
pandas
numpy
matplotlib
scikit-learn

---

🌐 Streamlit Application

The application allows users to:

• Enter YouTube video metrics
• Select category, device, country, and month
• Predict expected ad revenue
• View business insights
• Explore model performance visualizations

---

🎯 Project Outcomes

✔ Cleaned and processed YouTube analytics data

✔ Built and evaluated five regression models

✔ Achieved 95.19% prediction accuracy (R² Score)

✔ Identified key drivers of YouTube ad revenue

✔ Developed an interactive Streamlit application

✔ Generated actionable business insights

---

🔮 Future Enhancements

• Integration with YouTube Analytics API
• Real-time revenue forecasting
• Advanced models such as XGBoost and LightGBM
• Cloud deployment
• Personalized creator recommendations

---