Skip to content

Data Sources

Principle

Only publicly available data sources and documents are allowed. Teams may use the examples listed below or select other suitable public sources. What matters is that origin, access path, and retrieval date are documented cleanly.

Mandatory minimum requirement

Every source used must be documented at least with URL, format, access path, and retrieval date. Official APIs or official downloads should be preferred over scraping.

Example sources

Source Type Typical access Example use
opendata.swiss / CKAN API open government data portal CKAN API, CSV, JSON, files cross-domain administrative and specialist data
BFS / data.bfs.admin.ch / PX-Web API statistics PX-Web API, JSON, tables demography, economy, society, geography
geo.admin.ch / technical services geodata and geoservices WMS, WMTS, vector tiles, downloads, technical documentation mapping context, layers, georeferencing
MeteoSwiss Open Data / STAC weather and climate data STAC, raster, measurement series, downloads climate, weather trends, exposure
Swiss Parliament Open Data / web services politics and parliamentary activity web services, structured queries motions, debates, political signals
Fedlex SPARQL law and regulation SPARQL endpoint regulatory change, legal context, amendments
LINDAS SPARQL linked open data SPARQL endpoint linking and semantic references
Swiss Federal Office of Energy energy and supply statistics official statistics, downloads, some APIs energy supply, consumption, infrastructure indicators

Selection guidance

  • Both structured and unstructured public sources may be used.
  • Official sources should be preferred over unofficial mirrors or replicas.
  • If documents, news, or studies are included, their origin should remain clear and durably referenceable.
  • Sources should fit together thematically and support meaningful indicators or weak signals.
Field Description
Source name of the organization or platform
URL unique link to the source or dataset
Format e.g. API, CSV, JSON, GeoJSON, PDF, SPARQL, STAC
Access path concrete endpoint, download path, or query route
Retrieval date date of the actual data retrieval
License or terms of use document when available
Geographic reference Switzerland, canton, municipality, raster, point data, etc.
Update logic static, periodic, near real-time, or unknown

Recommendation for teams

Start with a small, well-documented source mix. A strong setup often combines:

  • one or two structured official datasets
  • at least one spatial source
  • at least one source for textual or regulatory signals

That creates a robust MVP with a visible evidence chain instead of an overly broad integration effort.