RAG System for Exam Generation

An AI-powered platform to create dynamic and adaptive exams from educational content

🚀 Project Overview

The RAG System for Exam Generation is a state-of-the-art platform designed to simplify the process of creating and managing exams. By leveraging Retrieval-Augmented Generation (RAG), the system retrieves relevant course content and uses advanced NLP techniques to generate questions dynamically. The project also includes features like automated grading and a user-friendly interface, ensuring a seamless experience for both educators and students.

---

📂 Features

• Knowledge Base Construction:
  Curated and structured database of course materials, including lecture notes, textbooks, and multimedia.

• Document Retrieval & Preprocessing:
  Efficient retrieval mechanisms and data preprocessing to extract relevant information.

• Dynamic Question Generation:
  AI-powered generation of multiple-choice and open-ended questions with customizable difficulty levels.

• Automated Answer Verification & Grading:
  Instant evaluation of responses with intelligent grading algorithms.

• Interactive User Interface:
  Built with Next.js, the platform offers a sleek, responsive interface for taking exams and receiving feedback.

---

🛠️ Tech Stack

Frontend

• Next.js: For creating a modern, responsive user interface.

Backend

• Python: Core language for retrieval, question generation, and grading logic.
• FastAPI/Flask: For building APIs to connect the front end with backend services.

AI & NLP Tools

• Hugging Face Transformers: For leveraging models like T5 and GPT for question generation.
• Elasticsearch: For efficient document retrieval.
• Sentence-BERT: For semantic similarity in answer grading.

Database

• PostgreSQL: To store and manage structured data.

---

📚 Project Structure

RAG-System-Exam-Generation/
├── frontend/            # Next.js frontend code  
├── backend/             # APIs, NLP pipelines, and grading logic  
├── data/                # Knowledge base data (lecture notes, textbooks, etc.)  
├── models/              # Pre-trained and fine-tuned NLP models  
├── scripts/             # Utility scripts for data preprocessing and testing  
└── README.md            # Project documentation  

---

🌟 How It Works

1. Knowledge Base Construction:
   Upload structured course materials into the system.

2. Document Retrieval:
   The RAG system retrieves relevant content using semantic search.

3. Question Generation:
   AI models generate exam questions tailored to the retrieved content.

4. Answer Verification & Grading:
   Student responses are automatically graded based on predefined algorithms.

5. User Interaction:
   Students access exams via a user-friendly interface and receive instant feedback.

---

🤝 Team Members

• Nourdin: Knowledge Base Construction
• Nordin & Youssef: Document Retrieval & Preprocessing
• Youssef: NLP for Question Generation
• Abdelfattah Bouhlali: Answer Verification, Grading, and User Interface

---

💻 Installation & Setup

1. Clone the repository:

   git clone https://github.com/<your-username>/RAG-System-Exam-Generation.git
   cd RAG-System-Exam-Generation

2. Install backend dependencies:

   pip install -r backend/requirements.txt

3. Run the backend server:

   python backend/app.py

4. Start the Next.js frontend:

   cd frontend  
   npm install  
   npm run dev  

5. Access the platform at http://localhost:3000.

---

🎯 Future Plans

• Expand question types (e.g., drag-and-drop, matching).
• Add support for adaptive exams based on student performance.
• Implement multilingual support for diverse educational contexts.

---

📜 License

This project is licensed under the MIT License. See the LICENSE file for details.