Queria

A geocoder for Japanese addresses

Turn an address into latitude and longitude, or click the map to find the address there. A CSV with an address column can be converted in one go. Free, no sign-up needed.

Area codes

Each address comes back with codes for joining statistics, along with latitude and longitude. You don't need to overlay boundaries to work them out.

  • Municipality code (local government code): join statistics by municipality
  • Census small area code: join population and households by chome
  • Grid square code (1km, 500m, 250m): join grid statistics
  • Postal code, and the town, block and building IDs of the Address Base Registry

Queria's datasets use the same codes, so converted addresses join directly with population and land price tables. When only the town's representative point is known, the small area and grid square come from that point too, so they can differ from the actual place in a large town.

Examples

  • Turn a list of addresses into coordinates

    Adds latitude, longitude and area code columns next to the address column of a CSV. The result loads straight into GIS software.

  • Normalize how addresses are written

    Writing variations such as “1-3” and “1丁目3番” come back in the notation of the national address data.

  • Look up the address and codes of a point

    From a position taken on site or from a photo, get the address, municipality code and small area code.

How it works

The Digital Agency's Address Base Registry (Japan's national address data) is imported every month into tables for address lookup. An incoming address is first read for its prefecture and municipality, and then only that municipality's town and number tables are read to match it. The answer is the representative point of the matched town or number.

Writing variations such as “1-3” versus “1丁目3番”, or old and new kanji forms, are normalized before matching. For a point on the map, the nearest address point is returned.

About the data

  • Built on the national address data

    The Digital Agency's Address Base Registry is imported every month. Every answer comes with its sources and the date of the data.

  • Matching rules and code are open

    The rules for normalizing and matching addresses are based on abr-geocoder, published by the Digital Agency. The code is open so anyone can check it.

  • Uploaded files are not sent

    The file is read inside this page, and only the address column is sent for conversion. Sent addresses are not stored.

Source code

Accuracy

  • Down to block and building number

    For addresses in the residential address system (○丁目○番○号), the point near the building is returned.

  • Down to town or chome

    When the number is unknown or has no point, the representative point of the town or chome is returned. It can be off by a few hundred meters.

  • Lot-number addresses

    Currently matched down to the town. Points for each lot are in preparation.

For a point on the map, the address of the nearest point is returned. Near town borders it can be the neighboring town.

Pricing

Free for now (beta). Up to 300 requests per minute from one place, and up to 1,000 rows per file conversion. If you need more, please contact us.

For developers

The same features are available as an API, with no sign-up or key. Fields are described in the API reference, and the code is on GitHub.

curl -s -G https://geocoder.queria.io/v1/geocode \
  --data-urlencode "address=東京都千代田区紀尾井町1-3"

Sources

When you publish results, include the sources attached to each answer.

  • Processed from the Address Base Registry (Digital Agency)
  • Processed from the Census boundary data (small areas) (Statistics Bureau, e-Stat)
  • Processed from the National Land Numerical Information (administrative areas) (MLIT)
  • Address matching rules: based on abr-geocoder (Digital Agency, MIT License)

During the beta the shape of the answers may change