GeoZones

Simple geospatial and administrative repository.

This dataset is built from the Official Geographic Code of INSEE, available via their [SparQL interface] (https://rdf.insee.fr/sparql/).

Model

There are two types of objects:

• levels
• areas

Zones

The file Zones {year} (json) is constructed from data extracted from the COG and contains, for all geographical scales, the following information:
- uri: Entity URI in INSEE RDF graph (example: "http://id.insee.fr/geo/arrondissement/6eeefa75-7352-48ee-884f-64783b8ca290"),
- name: name of the entity (example: "Lyon"),
- INSEE code: INSEE code of the entity (example: "691"),
- nameWithoutArticle: name without article of the entity (example: "Lyon"),
- codeArticle: Entity item code (example: "0"),
- type: type of entity (example: "Arrondissement"),
- is_deleted: boolean indicating whether the entity has been administratively deleted (example: true),
- level: level of scale of the entity (example: "fr:arrondissement"),
- _id: full identifier used by data.gouv.fr (example: "fr:arrondissement:691")

The Countries only zones {year} (json) file is a sample of the global Countries only zones {year} (json) file which contains only the countries.

Levels/Levels

The file contains the different possible scale levels, with the following information:
- id: entity level of scale, which corresponds to the ‘level’ field in the Zones file (example: "en:region"),
- label: naming the scale level (example: "French region"),
- admin_level: Scale level code (example: 40),
- parents: directly higher level(s) of scale (example: ["country"])

Construction

This dataset is built from the INSEE COG via a python script available [here] (https://github.com/etalab/datagouvfr_data_pipelines/tree/main/data_processing/geozones).

History

• 30/04/2015: first version
• 15/04/2016: addition of the URLs of the coats of arms/flags and an export using msgpack in order to reduce the size of the generated archive
• 19/04/2016: correction version providing a finer cut of the shapes of the municipalities
• 09/06/2016: correction version adding the parents for the municipalities of Corsica/DROM-COM and calculating the population for the districts
• 15/06/2017: version including data from GeoHisto and using GeoIDs, integrates 2017 data (COG, OSM).
• 28/08/2017: Added EPCI history from GeoHisto.
• 08/05/2019: Switching to COG 2019, bug fixing, adding the "geonames" key, switching to Wikidata, cantons and iris are no longer exported
• 30/11/2023: The data comes from the INSEE COG from their SparQL interface

Archives

Levels/Levels

They make it possible to model the different known levels of the referential and their theoretical relationships.
Their name is translatable.

Zones

A zone is the association of a unique identifier with a geographical polygon, a level and a name.
It has less than one unique code for the level.
It may have several known identifiers, which are not necessarily unique.

The name is optionally translatable (ex: European Union, World)

The following attributes are exported to the GeoJSON:

• id: A unique identifier following the specification GeoID
• code: The unique identifier for a given date of the zone for its level
• level: The identifier of the level of attachment
• name: The display name of the area in English (may be translated)
• population: Approximate/estimated population (optional)
• area: Estimated/approximate area in km2 (optional)
• wikidata: The associated Wikidata node (optional)
• wikipedia: A reference to Wikipedia (optional)
• dbpedia: A reference to DBPedia (optional)
• flag: A reference to the DBPedia flag (optional)
• blazon: A reference to the DBPedia blazon (optional)
• keys: a dictionary of the different codes known for this area
• parents: an unordered list of the identifiers of the different known parents
• ancestors: the list of possible ancestors
• successful: the list of possible successors
• validity: a period of validity (object with the attributes ‘start’/‘end’) (optional)

Construction

This dataset is built with the tool GeoZones
whose code is published on Github.

You can find the [detail of French specificities] (https://github.com/etalab/geozones/blob/master/LISEZMOI.md) on the repository.

Possible future improvements

Fields

• Overall weight = f(population, area, level)

Deliverables

• Various clarifications
• Localized JSON (in English only for now)
• Translations in JSON (as a hard alternative to the current PO/MO format)
• Level statistics (number of zones, coverage of attributes...)