Youness331/Darija-sentiment-analysis-public

# 📊 Darija Sentiment Analysis Project A comprehensive sentiment analysis application for Arabic/Darija text that combines web scraping, machine learning, and deep learning to analyze public sentiment from news articles and comments. ## 🎯 Project Overview This project provides an end-to-end solution for sentiment analysis of Arabic/Darija text, featuring: - **Web scraping** from Hespress news website - **Traditional machine learning** models (SVM, KNN, Naive Bayes) - **Deep learning** models (Bidirectional LSTM, CNN) - **Interactive web interface** for real-time analysis - **Comprehensive text preprocessing** for Arabic/Darija text - **Visualization tools** including word clouds and sentiment charts ## 📁 Project Structure ``` darija-sentiment-analysis/ ├── app.py # Main Flask application ├── package.py # Core scraping and preprocessing functions ├── requirements.txt # Python dependencies ├── .env # Environment variables (API keys) ├── .gitignore # Git ignore rules ├── .gitattributes # Git LFS configuration ├── README.md # Project documentation ├── sentiments analysis/ │ └── ml.ipynb # Machine learning notebook ├── models/ │ └── model/ │ └── model.safetensors # Pre-trained DarijaBERT model ├── templates/ # HTML templates for web interface │ ├── index.html # Homepage │ ├── results.html # Search results page │ ├── comments.html # Individual article comments │ └── all_comments.html # Aggregate sentiment analysis ├── static/ # Static files (CSS, JS, images) └── Stop_words.csv # Darija/Arabic stopwords dataset ``` ## 📓 Notebooks Overview ### `sentiments analysis/ml.ipynb` A comprehensive Jupyter notebook that implements and compares multiple sentiment analysis approaches: ## …