🧠 AI Quiz Generator

An intelligent quiz generation web app powered by NLP Transformer models. Automatically generates quizzes from Wikipedia content using BERT (NER) and T5 (Question Generation) — no manual question writing needed!

Built as a Python project submission for B.Tech CSE AIML at DAVIET, PTU.

Screenshots

Homepage

Quiz Generated

End of Quiz

---

✨ Features

• 📖 Auto-fetches Wikipedia content by topic
• 🤖 BERT NER (dslim/bert-base-NER) identifies key entities from text
• ❓ T5 transformer (valhalla/t5-base-qg-hl) generates natural language questions
• 3 question types generated automatically:
  • 🔵 Multiple Choice Questions (MCQ) with 4 options
  • 🟢 One-word answer questions
  • 🟡 True / False questions
• 🎛️ Choose number of questions to generate
• ⚡ Cached model loading for fast performance
• 🌐 Streamlit web interface — runs in browser

🛠️ Tech Stack

Models: dslim/bert-base-NER · valhalla/t5-base-qg-hl
Libraries: transformers · wikipedia · streamlit · random

🧠 How It Works

User selects topic
       ↓
Wikipedia content fetched
       ↓
Text cleaned & split into sentences
       ↓
BERT NER → identifies key entities (people, dates, places, concepts)
       ↓
T5 QG model → generates questions from highlighted entities
       ↓
Quiz displayed on Streamlit UI (MCQ + One-word + True/False)

📚 Supported Topics

Python · JavaScript · C · C++ · Rust · Java

🚀 How to Run

# Clone the repo
git clone https://github.com/Simran-775/Quiz-Generator.git
cd Quiz-Generator

# Install dependencies
pip install transformers wikipedia streamlit

# Run the app
streamlit run streamlit_app.py

⚠️ First run will download the BERT and T5 models (~500MB). After that, models are cached for fast loading.

📁 Project Structure

Quiz-Generator/
├── streamlit_app.py   # Streamlit UI
├── project.py         # Quiz logic (NER + QG pipelines)
└── README.md

📬 Contact

Made with 💙 by Simranjeet Kaur