🎬 Age Rating Prediction for Streaming Platforms using Naive Bayes

A machine learning project that predicts age ratings (e.g., PG, R, TV-MA) for movies and TV shows using metadata and content descriptions from major streaming platforms.

📌 Overview

This project applies Multinomial Naive Bayes, a text classification algorithm, to predict age ratings for shows and movies across Netflix, Disney+, Hulu, and Amazon Prime Video. By automating the classification based on descriptions, the model assists in improving viewer safety and content moderation.

🗂 Dataset

Source: Kaggle

• Netflix
• Disney+
• Amazon Prime
• Hulu

📌 These datasets contain:

• title, type, description, cast, genre, release_year, duration, rating, and more
• The rating column is the target variable

Note: Raw data files are excluded due to size. See data/README.md for download instructions.

🧠 Methodology

Preprocessing

• Missing values handled
• Text normalization (lowercasing, punctuation removal, stopword removal)
• Feature selection focused on description and other key metadata

Feature Extraction

• CountVectorizer and TF-IDF used to transform descriptions into feature vectors

Model

• MultinomialNB (from sklearn.naive_bayes)
• Trained on processed description data
• Tested using multiple evaluation metrics

📈 Evaluation Metrics

• Accuracy
• Precision
• Recall
• F1 Score
• Confusion Matrix

🛠 Tools Used

• Python
• Jupyter Notebook
• Pandas, NumPy
• Scikit-learn