YoussefNKH/tunisian-feedback-classifier

# 🧠 NLP - Detection of the Problem Source from Arabic Feedback This project uses Natural Language Processing (NLP) to classify Arabic customer feedback as either related to the **product** (المنتج) or the **service** (الخدمة). It includes a trained model, a Streamlit web app, and a Jupyter Notebook for training and evaluation. --- ## 📂 Project Structure ```plaintext tunisian-feedback-classifier/ │ ├── app.py # Streamlit app interface ├── NLP_Classification.ipynb # Notebook with full training & evaluation ├── MLP_model.pkl # Trained MLP neural network model ├── tfidf_vectorizer.pkl # TF-IDF vectorizer ├── finals.csv # Contains the data (Comment_Text_Arabic,Problem_Source(Labels)) ├── requirements.txt #Contains the requirements to make the interface work ├── README.md # Project documentation (this file) ``` --- ## 🚀 Features - Classifies Arabic feedback as either about the **product** or the **service** - Neural network (MLP) trained on TF-IDF features - Arabic-specific text preprocessing (normalization + stopword removal) - Interactive web app with Streamlit --- ## 🛠️ Installation ### 1. Clone the Repository ```bash git clone github.com cd tunisian-feedback-classifier ``` ### 2. Create Virtual Environment (Optional) ```bash python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate ``` ### 3. Install Dependencies ```bash pip install -r requirements.txt ``` --- ## ▶️ Run the Streamlit App ```bash streamlit run app.py ``` --- ## 📈 Model Performance | Model | Accuracy | F1-Score | |---------------|----------|----------| | Naive Bayes | 0.92 | 0.92 | | Neural Net ✅ | 0.95 | 0.95 | | RBF SVM | 0.95 | 0.95 | | Linear SVM | 0.93 | 0.93 | --- ## 👤 Author **Youssef Nakhli** 🎓 Data Engineer Student 📫 LinkedIn | GitHub