Business-intelligence classification of 49 traditional Turkish recipes along a diet–gut-barrier axis (SDG 2 → SDG 3, support SDG 12). Scores are an exploratory, transparent framework — not a validated index.
1 · Recipe typology — the classification core
Gut-support × Sustainability quadrant
Which recipes should the rewritten book champion, and which are the improvement targets?
Crosshairs = medians. Top-right = star recipes (high on both). Bottom-left = improvement opportunity, the recipes students target when rewriting the book. Filters above re-plot this chart.
2–5 · Overview (all 49 recipes)
2 · Region comparison
Is one region more gut-friendly?
3 · Category comparison
Why are desserts at the bottom and mezes at the top?
4 · Driver prevalence
How common is each gut-support driver?
Only 7 of 49 carry live probiotics — the two-pathway finding: live fermentation vs. legume/wild-green fiber.
5 · Dessert nuance
Are all desserts equal?
Syrup-drenched desserts sink; natural-sweetener desserts (mulberry pestil, grape molasses) score far better and are more sustainable.
6 · Regional signatures — the cultural bridge
Method · Weight sensitivity analysis
Do the conclusions depend on our arbitrary weights?
Re-scoring under 4 weight schemes to test robustness.
Method: gut-support recomputed as a weighted sum of the same binary/ordinal features under alternate weight vectors. If the star and bottom recipes stay stable, our narrative does not hinge on the exact weights — the key limitation reviewers will probe. Companion methods (not shown): cross-tabulation, and inter-rater reliability (Krippendorff's α) once students double-code.