iy3d1243/tunisian_NLP

# Arabic Complaint Classifier try it now : tunisiannlp-vft2qtepl3q4zqd… A machine learning application that classifies Arabic complaints using various classification models. ## Project Overview This application uses Natural Language Processing (NLP) techniques to classify Arabic complaints into different categories. It includes: - Text preprocessing for Arabic language - Multiple machine learning models: - Logistic Regression - Naive Bayes - Support Vector Machine (SVM) - Random Forest - A Streamlit web interface for easy interaction ## Files in the Project - `app.py`: Streamlit web application for the classifier - `models.py`: Script for training and saving the machine learning models - `*.pkl`: Saved models and vectorizer - `requirements.txt`: List of required Python packages ## How to Run 1. Install the required dependencies: ``` pip install -r requirements.txt ``` 2. Run the Streamlit app: ``` streamlit run app.py ``` 3. To retrain the models, run: ``` python models.py ``` ## Features - Arabic text preprocessing - Multiple classification models - Interactive web interface - Sample text generation - Model performance information ## Requirements - Python 3.6+ - Streamlit - Scikit-learn - Pandas - XGBoost