Sources and licenses
Where the material on this site comes from, and the terms it is used under.
Last updated 2026-09-07
Korean meanings and readings
The Korean meanings and readings for 9,210 characters are compiled and maintained by this project, built on the Korean Language Society's graded hanja list.
Example words
Words and their definitions come from Woorimalsam, published by the National Institute of Korean Language.
Woorimalsam is licensed CC BY-SA 2.0 KR, and the definitions shown here are available under the same terms.
Korean glosses for Japanese readings
Built by inverting the Japanese equivalents in the Basic Korean Dictionary (National Institute of Korean Language).
That dictionary is licensed CC BY-SA 2.0 KR, and the material shown here is available under the same terms.
Where the Basic Korean Dictionary has no matching entry, the gloss was generated by a large language model (Claude) from the EDRDG English definition and the character's Korean readings. These were not checked by a lexicographer; they cover roughly 73% of kun readings and 58% of Japanese words.
Machine-generated glosses are stored separately from the sourced data, so they can be replaced if better dictionary material becomes available.
Japanese data
School grades, newspaper frequency, kana readings and Japanese example words come from EDRDG's KANJIDIC2 and JMdict.
Both are licensed CC BY-SA 4.0, and that material is available here under the same terms.
English glosses, pinyin, strokes, radicals
Taken from the Unicode Consortium's Unihan database, used under the Unicode Terms of Use.
Character components
Glyph decomposition comes from KanjiVG, copyright Ulrich Apel, released under CC BY-SA 3.0. The corresponding content on this site is available under the same terms.
The one-line picture and use written for each of the 400 most common components were written for this site.
Most Han characters pair a meaning part with a sound part. Where a part only carries sound, that is what the note says; where the glyph origin is disputed, the note says so rather than inventing one.
How a character came to be written that way
The one-line glyph origin notes are Korean renderings of the Glyph origin sections of the English Wiktionary, which is available under CC BY-SA 4.0. The corresponding content on this site is available under the same terms.
The rendering was done by a large language model (Claude), given the source text and told to write only what it contained, and to say so where the origin is disputed. It was not checked by a lexicographer.
Glyph etymology is contested in many places. Consult the original sources where a scholarly judgement matters.
Phonetic series
Which characters share a sound part comes from Unihan's kPhonetic field, based on G. Hugh Casey's phonetic index.
That field does not say which part carries the sound, so it was worked out from the stroke count left after removing the radical, then checked against the phonetic parts KanjiVG marks by hand. Characters where the two disagree are left out.
Typeface
Chinese characters are set in LXGW WenKai TC, distributed under the SIL Open Font License 1.1. The full licence is served alongside the font.
A note on accuracy
Errors can occur when data is compiled. For scholarly or legal purposes, please check the original sources.