bas07ha/darija-sentiment-analysis

Sentiments classification of Darija text using NLP and machine learning techniques # Darija Sentiment Analysis This project performs sentiment classification on Darija (Algerian Arabic dialect) text using Natural Language Processing and machine learning techniques. ## Objective The goal is to classify text comments into positive or negative sentiment based on their content. ## Dataset The dataset contains 1500 labeled Darija comments used for training and evaluation. Train/Test split: - 80% training - 20% testing ## Methodology The workflow includes: 1. Data loading and preprocessing 2. Text cleaning and normalization 3. Feature extraction 4. Model training 5. Model evaluation ## Evaluation Metrics The models are evaluated using: - Precision - Recall - F1-score - Support - Confusion Matrix ## Visualizations The project includes several visualizations: - Confusion matrix - Word frequency histograms - Most frequent words in positive sentiment - Most frequent words in negative sentiment - Word distribution for training and test datasets ## Technologies - Python - Pandas - Scikit-learn - Matplotlib - Jupyter Notebook ## Project Structure notebook/ : sentiment analysis code data/ : dataset used for training and testing images/ : visualizations generated from the analysis