📩 AI Spam Detector (Streamlit App)

A simple and effective AI-powered web app built using Streamlit that detects spam or fraudulent SMS messages using Natural Language Processing (NLP) and Machine Learning.

This project supports:

• ✅ Manual SMS input
• ✅ Bulk spam detection via .csv or .pdf upload
• ✅ Downloadable prediction results
• ✅ Model persistence using joblib

---

🚀 Live Demo

👉 Click to Use the App

---

📂 Folder Structure

AI-Spam-Detector/ ├── app.py # Streamlit web app ├── spam_model.pkl # Trained spam detection model ├── vectorizer.pkl # TF-IDF vectorizer ├── requirements.txt # Python dependencies ├── README.md # Project documentation └── spam.csv # Dataset used for training (optional)

---

📥 Features

• 🔍 Single Message Detection
  Type any message and instantly check if it's spam or not.

• 📂 Bulk Detection
  Upload a .csv or .pdf file of messages. The app processes each one and tells you if it's spam.

• 📤 Download Results
  Export prediction results as CSV.

---

📊 Technologies Used

• Python 3
• Streamlit
• scikit-learn
• Pandas
• Joblib
• PyMuPDF (fitz for PDF reading)
• Natural Language Toolkit (NLTK)

---

⚙️ Run Locally

1. Clone the repo:

   git clone https://github.com/prangya0312/ai-spam-detector.git
   cd ai-spam-detector

pip install -r requirements.txt

streamlit run app.py

🧠 Model Details Algorithm: Multinomial Naive Bayes

Trained on: SMS Spam Collection Dataset

Text Processing: TF-IDF vectorization

🛡️ Future Enhancements Add login/authentication

Store analysis history per user

REST API version

Mobile UI optimization

📜 License This project is open-source and available under the MIT License.

🙋‍♀️ Developed By Prangya Gantayat B.Tech (CSE) | SOA University 📍 Odisha, India