Corpus composition research
Kanji Pair Research
Enter one kanji to discover other characters recorded in the same JMnedict name spellings, or enter two kanji to inspect that exact co-occurrence. Every count in this research view describes the imported corpus—not population popularity in Japan.
香
Jōyō · KANJIDIC grade 4 · 9 strokes
留incense, smell, perfume
Jōyō · KANJIDIC grade 5 · 10 strokes
Find exact-kanji names →detain, fasten, halt
Recorded names
Name readings containing both characters
The pair only has to occur in the same written form; the two kanji are not required to be adjacent unless counted in the adjacency statistics above.
Given name
晴留香
5 source-recorded spellings in this reading group contain both 香 and 留Given nameHaruka
はるか
日香留
3 source-recorded spellings in this reading group contain both 香 and 留Given nameHikaru
ひかる
留実香
2 source-recorded spellings in this reading group contain both 香 and 留Given nameRumika
るみか
実香留
2 source-recorded spellings in this reading group contain both 香 and 留Given nameMikaru
みかる
香留奈
2 source-recorded spellings in this reading group contain both 香 and 留Given nameKaruna
かるな
実留香
2 source-recorded spellings in this reading group contain both 香 and 留Given nameMiruka
みるか
留香
1 source-recorded spelling in this reading group contain both 香 and 留Given nameRuka
るか
香留
1 source-recorded spelling in this reading group contain both 香 and 留Given nameKaoru
かおる
留香子
1 source-recorded spelling in this reading group contain both 香 and 留Given nameRukako
るかこ
香奈留
1 source-recorded spelling in this reading group contain both 香 and 留Given nameKanaru
かなる
香留鼓
1 source-recorded spelling in this reading group contain both 香 and 留Given nameKaoruko
かおるこ
香穂留
1 source-recorded spelling in this reading group contain both 香 and 留Given nameKahoru
かほる
衣留香
1 source-recorded spelling in this reading group contain both 香 and 留Iruka
いるか
Evidence boundary
Co-occurrence is not popularity or etymology
The pair relation comes from exact JMnedict source rows through the compact kanji-usage index. KANJIDIC2 contributes character metadata only. A high corpus count can reflect dictionary coverage and should not be read as a national usage frequency.