🤖 Rule-Based Chatbot

A rule-based chatbot built in Python — a terminal version and a desktop GUI, both sharing the same logic engine. Understands greetings, jokes, riddles (with a real answer-checking game), study/work motivation, simple math, and basic Roman Urdu/Hinglish phrasing.

---

✨ Features

• Keyword + typo-tolerant matching (so "helo" still works)
• Riddle mini-game — asks a question, checks your guess, only reveals the answer if you say you don't know
• Motivation quotes that adapt to context ("motivate me for study" vs "motivate me for work")
• Understands common Roman Urdu / Hinglish phrases and replies in kind
• Remembers your name mid-conversation
• Simple arithmetic ("what is 12 + 7")
• Desktop GUI with dark/light mode, quick-actions sidebar, session analytics, and chat export
• Data-driven — knowledge base lives in intents.json, not hardcoded, so adding a topic doesn't require touching code

---

🧠 How It Decides What to Say

For each message: clean up text → check for an exact match → check if a riddle answer is pending → check for a "tell me another one" follow-up → check for a name introduction → check for math → scan for topic keywords → fall back to a default reply if nothing matched.

---

🛠️ Tech Stack

Layer	Technology
Language	Python
GUI	Tkinter
Testing	unittest

---

🚀 Getting Started

git clone https://github.com/areebaathar-dev/rule-based-chatbot.git
cd rule-based-chatbot
pip install -r requirements.txt

python gui_app.py        # GUI version
python cli_chatbot.py    # Terminal version

Running the tests

python -m unittest discover -s tests -v

---

📁 Project Structure

rule-based-chatbot/ ├── chatbot_logic.py # Core logic (used by both interfaces) ├── cli_chatbot.py # Terminal version ├── gui_app.py # Desktop GUI ├── intents.json # Knowledge base (keywords + responses) ├── benchmark.py # Dictionary lookup vs if-elif comparison ├── tests/ │ └── test_chatbot_logic.py # Unit tests └── requirements.txt

---

🔭 Extending It

Add a new topic by editing intents.json — no code changes needed:

"favorite_color": {
  "keywords": ["favorite color", "favourite colour"],
  "responses": { "en": ["I'd say blue, what's yours?"] }
}

📄 License

MIT — see LICENSE.

---

👩‍💻 Author

Areeba Athar BS Computer Science, NFC-IEFR Faisalabad LinkedIn · GitHub