GSoC 2024 Proposal: Sorting Hat AI Math Game
for Sugar Labs
Author: Baati Yosri
Email: [email protected] | GitHub: Baati-Yossri | LinkedIn: yosri-baati
Timezone: UTC+1 | First Language:Arabic
1. Basic Details
While I’m new to open-source contributions, I’m deeply motivated to learn and
contribute meaningfully to Sugar Labs. Here’s how I’ve prepared:
● Technical Readiness: I’ve practiced Python by completing an NVIDIA course:
Getting started with deep learning.
● Passion for Education: I’ve always been fascinated by the intersection of math
and games—whether it’s my daily Sudoku practice or exploring puzzles like the
Fifteen Puzzle and Soma cubes.
I’ll prioritize transparency with mentors, ask questions early, and document my
progress to ensure a smooth onboarding.
2. Project Details
What am I making?
A Sorting Hat AI game where kids teach an AI to classify objects
(animals/shapes/numbers) using Decision Trees/k-NN. Correcting the AI’s mistakes
makes it "learn" over time, introducing ML concepts playfully.
Impact on Sugar Labs:
● Adds an interactive, AI-themed math game to Sugar’s educational toolkit.
● Teaches critical thinking and pattern recognition through hands-on ML.
● Complements Sugar’s goal of "learning by doing" for children.
Technologies:
● Python (core logic), GTK/Sugar Toolkit (UI), scikit-learn (k-NN/Decision Trees),
PyGame (optional for visuals).
3. Timeline
Weeks Tasks Deliverables
1–2 Research k-NN/Decision Trees; design UI Approved design doc;
mockups; scaffold Sugar activity. basic activity structure.
3–5 Implement core classification logic; Functional prototype
integrate simple datasets (Evaluation 1).
(animals/shapes).
6–8 Add "teaching" mechanics (user AI learns from user input.
corrections improve AI).
9–10 Polish UI/UX; write tests; document Game ready for testing
code. (Evaluation 2).
11–12 Mentor feedback, bug fixes, final Merged code; demo
submission. video.
● Hours/Week: 30 hours (consistent with 350-hour total).
● Unavailable July 27 due to prior commitment (off-grid).
● Progress Reports: Biweekly blog posts + GitHub updates.
4. Post-GSoC Plans
"I’ll maintain the game post-GSoC, add more datasets (e.g., numbers), and mentor future
contributors. Long-term, I aim to expand Sugar’s AI-education toolkit."
