Attribution & Credits
Our rule: nothing appears in Michi without its creator named. Every video, article, dataset and tool we point you to is credited at the point of use — not buried on a page you'll never visit.
This file is the master record. The same information is rendered in-app on the Credits page, and a short credit line appears directly beneath every embedded or linked resource.
1. How we use third-party material
We are deliberately conservative. There are only three things we do:
| What we do | Example | Why it's OK |
|---|---|---|
| Link out to a creator's own page | "Read Tofugu's hiragana guide →" | Linking is not copying. Traffic goes to the creator. |
| Embed via the platform's own player | A YouTube lesson in an <iframe> |
YouTube's official embed player is designed for this, serves the creator's ads/analytics, and links back to their channel. |
| Bundle openly-licensed data | Dictionary entries under CC BY-SA | Explicitly permitted by the licence, with the required attribution. |
What we never do:
- ❌ Re-host or mirror someone's video, audio, PDF or article
- ❌ Copy explanations, mnemonics or example sentences from textbooks, paid courses, or other learning sites
- ❌ Reproduce official JLPT past papers or copyrighted exam questions
- ❌ Strip branding, bypass paywalls, or present someone else's work as ours
All explanations, mnemonics and example sentences in content/ are original writing created for this project.
2. Attribution format
Every resource carries a credit line rendered directly below it:
Video by {creator} · watch on {platform} → {url}
Data from {dataset} by {maintainer}, used under {licence} → {licence url}
The data behind these lines lives in each track's resources.json, so a credit can never be accidentally separated from the thing it credits.
3. Open datasets we rely on
These are the standard, openly-licensed sources for Japanese language data. Each entry's exact licence is verified against the source project before use — see docs/_research/RESEARCH_NOTES.md and the machine-readable resources.json in each track.
| Source | Maintainer | Typical licence | Used for |
|---|---|---|---|
| JMdict / EDICT | Electronic Dictionary Research and Development Group (EDRDG), Monash University | CC BY-SA | Dictionary definitions and readings |
| KANJIDIC2 | EDRDG, Monash University | CC BY-SA | Kanji readings, meanings, stroke counts |
| KanjiVG | Ulrich Apel and contributors | CC BY-SA | Stroke-order diagrams |
| Tatoeba | Tatoeba Project contributors | CC BY | Example sentences (where used) |
ShareAlike note. Several of these are ShareAlike licences. That is exactly why our original content is released under CC BY-SA 4.0 — it keeps us compatible when derived material is involved, and it means anything built on our content stays open too.
4. Creators we send you to
The full, current list lives in content/tracks/<track>/resources.json and on the in-app Credits page. It includes independent teachers on YouTube, free reference sites, and free tools.
If you are one of these creators: thank you. Your work is the reason a free roadmap like this is possible.
5. Official JLPT trademarks
Michi is an independent, unofficial study aid. It is not affiliated with, endorsed by, sponsored by, or connected to the Japan Foundation, Japan Educational Exchanges and Services (JEES), or any official JLPT body.
"JLPT" and "Japanese-Language Proficiency Test" are used descriptively, to identify the exam this track prepares you for — a nominative fair use. All trademarks belong to their respective owners. We do not reproduce official exam materials; for official information and sample questions, always go to the organisers' own site.
6. Takedown requests
If you are a rights holder and believe something here uses your work improperly — or you'd simply prefer we link differently, credit differently, or not at all — we will act quickly and without argument.
Please include the URL or file path, what the material is, and what you'd like done. We aim to respond within 72 hours. We would rather remove something instantly than have a creator feel poorly treated.
7. Contributing attributions
Adding a resource to a track? It must include: title, creator, url, license, attribution, and an honest embeddable flag. Pull requests that add a resource without full attribution will be asked for changes. See CONTRIBUTING.md.