LearnMate AI Chatbot

Overview

LearnMate is an AI-powered educational chatbot designed to assist students in understanding academic materials more effectively. The system uses a Retrieval-Augmented Generation (RAG) approach to provide accurate, context-based answers from course-related documents.

Unlike traditional chatbots, LearnMate retrieves relevant information from a knowledge base before generating responses, ensuring higher accuracy and relevance.

---

Features

• 📚 Context-aware chatbot using RAG architecture
• 🔍 Retrieval of relevant content from academic materials
• 💡 Follow-up prompt suggestions for continuous learning
• 🧠 Topic-based learning support
• ⚡ Fast response using vector search (FAISS)

---

Tech Stack

• Python
• FAISS (Vector Database)
• LLM Integration (Ollama)
• VS Code

---

System Architecture

The system follows a Retrieval-Augmented Generation pipeline:

1. Data Preprocessing

   • Convert documents (PDF, PPTX, DOCX) into text
   • Split text into smaller chunks

2. Embedding Generation

   • Convert text chunks into vector embeddings

3. Vector Storage

   • Store embeddings using FAISS index

4. Query Processing

   • User inputs a question

5. Retrieval

   • System retrieves most relevant chunks

6. Response Generation

   • LLM generates answer based on retrieved context

---

Project Structure

LearnMate_Project/
│
├── app/                # Main application (chatbot interface)
├── rag/                # Retrieval and chatbot logic
├── preprocessing/      # Text extraction and chunking
├── embeddings/         # Embedding generation scripts
├── vectorstore/        # FAISS index handling
│
├── .gitignore
└── README.md

---

How to Run

1. Clone the repository:

git clone https://github.com/Edryana/LearnMate_Project.git
cd LearnMate_Project

2. Install dependencies:

pip install -r requirements.txt

3. Run preprocessing (if needed):

python preprocessing/convert_to_text.py
python preprocessing/chunk_text.py

4. Generate embeddings:

python embeddings/create_embeddings.py

5. Run the chatbot:

python app/app.py

---

Notes

• Dataset files are not included due to size limitations
• Users may need to provide their own academic materials for testing

---

Future Improvements

• Web-based UI integration
• More advanced ranking for retrieval
• Support for multiple subjects
• Performance optimization

---

Author

Edryana Bachelor of Computer Science (Hons.) UiTM Jasin, Melaka