Japanese Address Normalizer (住所正規化)

Turn messy Japanese addresses into clean, structured data with coordinates.

• Unifies notation variants: 丸ノ内 → 丸の内, 1丁目 → 一丁目, full-width digits, ヶ/が, old/new kanji
• Fills in a missing prefecture or county (千代田区… → 東京都千代田区…)
• Splits each address into prefecture / city / town (町字・丁目) plus the remaining block number
• Adds latitude / longitude of the town and official codes (local government code, ABR 町字ID)
• Runs entirely on bundled open data, so your addresses never leave the run and there are no third-party API limits

Typical uses: deduplicating customer lists (名寄せ), cleaning CRM/EC data, joining address lists with statistics, plotting customers on a map.

Input

{ "addresses": ["東京都千代田区丸ノ内1丁目1-1", "千代田区霞ヶ関１－２", "大阪市北区梅田3-1-3"] }

Output (one row per address)

field	example	meaning
input	千代田区霞ヶ関１－２	the address as given
normalized	東京都千代田区霞が関一丁目2	full normalized string
prefecture / city / town	東京都 / 千代田区 / 霞が関一丁目	structured parts
rest	2	block / building part after the town
level	3	0 = not recognized, 1 = prefecture, 2 = city, 3 = town
lat / lng	35.6753 / 139.7524	representative point of the town
cityCode / machiazaId	131016 / 0002001	local government code and ABR town ID
error	null	reason when an address could not be processed

Pricing

You pay per successfully processed address, about $0.50 per 1,000 addresses. Empty or failed rows are free. The run stops automatically at your spending limit, so a test with a few addresses costs a fraction of a cent.

FAQ

Are my addresses sent anywhere? No. The address data is bundled inside the Actor, so nothing is sent to a third-party API and nothing is stored after the run except your own dataset.

How do I call it from code or an AI agent? Use the Apify API or the Apify MCP server with the input shown above; results come back as JSON or CSV from the run's dataset.

What if an address can't be matched to a town? You get the deepest level that matched (level 1 = prefecture, 2 = city), so you can filter those rows. Rows where not even the prefecture or city could be recognized get an error and are free.

Data sources and credits

• 「アドレス・ベース・レジストリ」（デジタル庁）
• 「位置参照情報」（国土交通省）
• Processed with Geolonia 住所データ (CC BY 4.0) and normalize-japanese-addresses (MIT) by Geolonia.

This Actor is not affiliated with or endorsed by the Digital Agency, MLIT or Geolonia. Results are best-effort and not guaranteed by any of them. The bundled data snapshot is refreshed when a newer release is published.

---

日本語の説明

表記ゆれのある日本の住所を、都道府県・市区町村・町字（丁目）に分けて正規化し、緯度経度と自治体コードを付けて返します。顧客名簿の名寄せや住所データの整理に使えます。住所データは Actor 内に同梱しているため、入力した住所が外部に送信されることはありません。