Sources and credits
Suwa draws on the work of dictionary editors, digitization volunteers, and open-source contributors.
JMdict / EDRDG
Dictionary content uses the English-language JMdict dataset. Copyright in the dictionary is held by James William Breen and the Electronic Dictionary Research and Development Group (EDRDG). The data is available under Creative Commons Attribution-ShareAlike 4.0.
Suwa imports JMdict data into relational tables and selects applicable readings and senses for dictionary lookup and study material. See the JMdict documentation and EDRDG dictionary license statement.
KanjiVG
Stroke data is adapted from KanjiVG, copyright © 2009–2026 Ulrich Apel and contributors, under Creative Commons Attribution-ShareAlike 3.0 Unported.
Suwa uses the July 14, 2026 release. Its adaptation retains ordered main-glyph stroke paths, converts path commands into absolute cubic curves, and removes stroke numbers, component metadata, and variant glyphs. The adapted stroke dataset remains under the same license. Full license text.
Aozora Bunko
Japanese source sentences are sampled from works digitized by Aozora Bunko and its volunteers. Suwa retains each selected work’s title, author, and source link, and displays source attribution with the challenge.
Study sentences omit ruby markup and source annotations while retaining the selected sentence’s wording. Aozora Bunko’s bibliographic metadata is provided under Creative Commons Attribution 4.0; literary works have their own copyright status. See Aozora Bunko’s file-use guidelines.
MeCab, Darts, and IPADIC
Suwa’s server-side Japanese analyzer is derived from MeCab 0.996 by Taku Kudo and Nippon Telegraph and Telephone Corporation, using its BSD license option. It includes code derived from Darts 0.31 and uses IPADIC dictionary data.
The full MeCab, Darts, IPADIC, and ICOT notices preserve the applicable copyright, permission, and warranty statements.
AI-generated study material
Reference translations, contextual explanations, difficulty estimates, and subscriber translation feedback are generated using OpenAI. Difficulty estimates draw on the model’s knowledge and guidance informed by JEV and the Hagoromo grammar database. Generated material can contain errors and is not an official proficiency assessment. See the AI processing disclosure.
Corrections and source questions
Report a source attribution, dictionary, or license concern to j.tan2231@gmail.com. Referencing a source does not imply its authors or maintainers endorse Suwa.