🎙️ Spotify Podcast Yapping Detection

A machine learning system that classifies Spotify podcasts as "yapping" (fast, informal speech) vs "normal" (structured, professional) content.

🚀 Quick Start

Prerequisites

• Python 3.7+
• Kaggle account with API credentials

Installation

git clone <your-repo>
cd spotify-yapping-detector
pip install -r requirements.txt

Usage

# 1. Process dataset
python scripts/spotify_processor.py

# 2. Train model
python scripts/train.py

📁 Project Structure

├── src/                    # Source code
├── scripts/               # Executable scripts
├── data/                  # Datasets
├── models/                # Trained models
├── images/                # Visualizations
├── notebooks/             # Jupyter notebooks
└── requirements.txt       # Dependencies

🎯 How It Works

The system identifies "yapping" podcasts using:

• Keywords: drama, react, gossip, tea, rant vs education, tutorial, news
• Language patterns: Multiple exclamations, ALL CAPS, long titles
• Content analysis: Reading complexity, sentence structure

📊 Results

The model achieves 99%+ accuracy with balanced datasets.

🔧 Customization

Add new keywords in src/yapping_detector.py:

self.yapping_keywords['custom'] = ['new', 'keywords']

📝 License

Educational and research use only.