DRC Administrative Hierarchy

Pipeline to build DR Congo’s administrative hierarchy (and roads) from an OSM .osm.pbf, store it in Postgres/PostGIS, and mirror it into Neo4j. The goal is to keep an updated database ready for applications and backend services that need administrative locations: address validation/lookup, populating select lists, or any feature that must link entities to a clean admin hierarchy.

Data sources and attribution,

• DR Congo OSM extract: https://download.geofabrik.de/africa/congo-democratic-republic-251204.osm.pbf
• OpenStreetMap CD (RDC) on HDX: https://data.humdata.org/organization/osm-rdc
• Upstream data © OpenStreetMap contributors. Big thanks to Claire Halleux for their contributions to OSM DRC: https://wiki.openstreetmap.org/wiki/User:Claire_Halleux.

What you get

• Up-to-date DR Congo admin hierarchy, ready for integration in apps/services that need address lookup, validation, and select lists.
• Domain tables for country → province → city/territory → municipality/commune → neighborhood/groupement → locality/village → bloc, with roads anchored to the deepest containing admin unit.
• Exports for Postgres and Neo4j (CSV + SQL) under dataset/.

Requirements

• Docker + docker compose
• osm2pgsql installed on the host
• OSM extract in data/ (default: data/congo-democratic-republic-251204.osm.pbf)

Setup

git clone https://github.com/bernard-ng/drc-admin-hierarchy drc-admin-hierarchy
cd drc-admin-hierarchy

# Place your .osm.pbf in data/ or set OSM_FILE when running the import
cp /path/to/your.osm.pbf data/congo-democratic-republic-251204.osm.pbf

Workflow

0. docker compose up settings in .env as needed.
1. make import — load OSM into Postgres.
2. make rebuild — build hierarchy + roads with admin links.
3. make graph — create Neo4j nodes/edges. (Neo4j: http://localhost:7474, Bolt: bolt://localhost:7687.)

Notes:

• Domain table values are trimmed and lowercased.
• Kinshasa is treated as both province and city (province mirrored as city when needed).
• .env is read by scripts and compose; adjust for custom ports/credentials.

Processed datasets

Processed outputs are stored under dataset/ after running make rebuild and make graph:

File	Description
data_admin.sql	SQL dump of admin hierarchy domain tables (country through bloc).
data_osm.sql	SQL dump of raw OSM tables (osm_boundaries, osm_places, osm_roads).
admin_nodes.csv	Neo4j-ready nodes for admin hierarchy.
admin_edges.csv	Neo4j-ready edges linking admin levels (parent/child).
roads.csv	Roads with names/refs and admin anchors.
road_admin_edges.csv	Edges between roads and their containing admin units.

Domain tables and admin levels

Domain tables align to OSM admin_level values (see https://wiki.openstreetmap.org/wiki/FR:Key:admin_level):

• country — level 2
• province — level 4
• city (territoire/ville, Kinshasa mirrored) — level 6
• municipality (collectivité/commune) — level 7
• neighborhood (groupement/quartier) — level 8
• locality (localité/village/cellule) — level 9
• bloc (subdivision interne) — level 10
• road — not an admin_level, but each road is anchored to the deepest admin unit it intersects (bloc → country).