AES Block Cipher Mode Selection Tool

This project helps users with no cryptography background select the right AES block cipher mode (ECB, CBC, OFB, CTR, GCM, SIV, XTS) by answering a short series of questions. It consists of two parts: a Python decision tree generator and a React web app.

Decision Tree Generator (decision_tree.py)

A standalone Python script that reads the implementation usage table (ImplementationUsageTable.csv) and builds a decision tree using the ID3 algorithm. The tree splits on security-relevant attributes (e.g. integrity, error propagation, parallelism) and uses weighted information gain with a cardinality penalty to produce a multi-question path rather than a flat one-shot tree.

Features:

• Interactive mode — asks multiple-choice questions in the terminal and recommends an AES mode
• "I don't know" support — skipping a question explores all branches and returns ranked results
• Evaluation — runs 1000 random trials to measure accuracy vs. a random-choice baseline
• JSON export — outputs the tree as quiz_tree.json for use by the React frontend

How to run

Make sure Python 3 is installed. No external dependencies are required.

# Interactive mode — answers questions to recommend an AES mode
python decision_tree.py

# Print the learned decision tree structure
python decision_tree.py --print-tree

# Run evaluation (1000 trials, accuracy vs random baseline)
python decision_tree.py --evaluate

# Export the tree as JSON for the React frontend
python decision_tree.py --export-json

React Web App (my-app/)

A React frontend that presents the decision tree as an interactive quiz with a visual, user-friendly interface. Users click through multiple-choice questions and receive a recommended AES mode with an explanation of why it fits their needs.

How to run

cd my-app
npm install
npm run dev

Integration

The decision tree generator and the React app are connected through the exported quiz_tree.json file. To update the quiz after changing the CSV table or tweaking the tree weights:

1. Edit ImplementationUsageTable.csv with new rows/columns as needed
2. Run the export:

   python decision_tree.py --export-json

3. Copy the output into the React app:

   cp quiz_tree.json my-app/src/quiz_tree.json

4. The React app reads this JSON to drive its quiz flow — no manual question editing needed.

This means the web app's quiz is always generated from the same ID3 algorithm and CSV data as the command-line version.