Open research asset · Batch 37
Japanese Baby Name Data
A reproducible export of reviewed Japanese given-name records, exact kanji-and-reading variants, literal glosses, and explicitly typed survey evidence. Choose JSON for nested data or CSV for one row per written variant.
Dataset coverage
- Reviewed names
- 150
- Exact variants
- 187
- Evidence-backed variants
- 185
- Ranking-backed variants
- 182
- Evidence rows
- 207
Schema v1 · generated 2026-07-30 · monthly evidence snapshot 2026-07
Field definitions
- Identity
- id, slug, romaji, hiragana, and katakana
- Pronunciation
- Approximate English guide and reviewed mora count
- Classification
- Gender labels and editorial usage status
- Exact variant
- Kanji, complete kana reading, and literal character glosses
- Evidence
- Source ID, year, sample rank, gender context, and evidence type
- Review trail
- Last-reviewed date and canonical detail URL
Evidence by source year
| Year | Evidence rows | Exact rankings |
|---|---|---|
| 2022 | 18 | 18 |
| 2023 | 14 | 14 |
| 2024 | 7 | 7 |
| 2025 | 165 | 152 |
Rows are observations, not unique people. The same reviewed variant may appear in more than one year or gender context.
Reproducibility and evidence types
Deterministic build
The export is generated from the publication-gated name records and three locked priority-evidence files. Re-running the repository generator rebuilds both public downloads from those inputs.
Typed observations
name-ranking is an exact written-name rank. Reading examples and cross-gender reading rankings remain separately labeled. An unranked documented example is never converted into frequency evidence.
Limitations to retain downstream
- The July 2026 snapshot is the dataset refresh month, not a July birth cohort or a live monthly popularity ranking.
- The latest ranking evidence is from 2025; the source is a named commercial survey sample, not a national census.
- Reading-ranking examples are labeled separately and must not be interpreted as full written-name positions.
- Documented-name examples establish an editorial example, not a ranking or frequency.
- Missing records or ranks mean the pair is not in this reviewed export; they do not prove that a name was absent from a source survey.
See the editorial methodology and human-readable ranking snapshot.