CS FAQ Chatbot

A modern AI powered FAQ chatbot designed to help Computer Science students by answering common CS related questions using semantic similarity.

This project was built as a learning focused AI application to understand how sentence embeddings and similarity based models work in real world scenarios.

---

🚀 Features

• ✅ Semantic question answering (not keyword-based)
• 🧠 Uses Sentence Embeddings + Cosine Similarity
• 🤝 Hybrid approach (Rule-based + Machine Learning)
• 🎨 Modern desktop UI (CustomTkinter)
• 🌍 English-only input validation
• ❓ Graceful handling of unknown questions
• 📦 Can be converted into a standalone .exe application

---

🛠️ Tech Stack

• Python
• Sentence Transformers
• PyTorch
• Pandas
• CustomTkinter (UI)

---

🧠 How It Works

1. Questions and answers are stored in a CSV file (faq.csv)
2. Questions are cleaned and converted into sentence embeddings using a pre-trained Transformer model
3. User input is embedded using the same model
4. Cosine similarity is used to find the most relevant question
5. The chatbot returns the corresponding answer if similarity is above a threshold
6. Simple rule-based logic handles greetings like "hello" or "hi"

---