Lightweight QA Bot

A CPU-friendly question-answering bot that learns from your Q&A pairs on the fly. Uses TF-IDF vectorization (instead of heavy neural models) + a C++ extension for blazing-fast cosine similarity search.

Features

• Self-learning — teaches itself new Q&A at runtime, persisted to ~/.minibot/qa.csv
• TF-IDF based — no large model downloads, minimal RAM, instant startup
• C++ accelerated — similarity search runs in native code via a Python C extension
• Potato-friendly — runs on any machine with Python 3.8+; no GPU, no heavy deps
• Modular — clean separation into src/storage.py, src/engine.py, and csrc/

Prerequisites

• Python 3.8+
• C++ compiler (g++ or clang) with Python dev headers
• pip install pandas scikit-learn

Build & Run

python setup.py build_ext --inplace
python main.py

Usage

python main.py

• Type a question; the bot replies from what it knows.
• If unsure (similarity < 0.4), it asks you to teach it.
• Type .exit to quit.

Structure

├── main.py              # CLI entry point
├── src/
│   ├── engine.py        # QA engine (TF-IDF + C++ search)
│   ├── storage.py       # CSV persistence layer
│   └── csearch.*.so     # compiled C++ extension
├── csrc/
│   ├── csearch.cpp      # Python C extension
│   └── csearch.h        # cosine similarity kernel
└── setup.py             # build script for C extension

License

MIT