Skosmos Vocabulary Extractor (SKOS thesaurus API)

Point this actor at any Skosmos server and pull structured SKOS vocabulary data — no per-service code required. Skosmos is the de-facto open-source vocabulary / thesaurus browser: it powers the FAO's AGROVOC, the ZBW's STW Thesaurus for Economics, France's Loterre, Finland's Finto, ACTRIS, and dozens of national and domain-specific vocabulary services. Every one exposes the same SKOS REST API — this actor speaks it generically.

SKOS (Simple Knowledge Organization System) is the W3C standard for thesauri, taxonomies, classification schemes, and subject-heading systems. If you build knowledge graphs, do data cleaning / subject indexing, or need controlled-vocabulary terms with their broader/narrower hierarchy and cross-vocabulary mappings, this is your extraction layer.

What it does

Five modes:

Mode	What it returns
vocabularies	Every vocabulary the server hosts — id, title, URI. The cheap connectivity probe; run it first to discover what's available.
search	Full-text concept search within a vocabulary (or globally across the whole server). One row per hit, carrying the prefLabel, altLabel, type, and language. Use a * suffix for prefix search.
data	Resolve a single concept URI to its full detail: every prefLabel in all languages, broader / narrower / related links, notation, definition, and cross-vocabulary mapping links (exactMatch / closeMatch).
topConcepts	The top-level concepts of a vocabulary — the roots of its hierarchy.
types	The RDF types (Concept, Collection, …) a vocabulary declares.

Example input

{
  "baseUrl": "https://agrovoc.fao.org/browse/rest/v1",
  "modes": ["vocabularies", "search"],
  "vocab": "agrovoc",
  "query": "water*",
  "lang": "en",
  "maxHits": 50
}

Resolve one concept to full detail:

{
  "baseUrl": "https://agrovoc.fao.org/browse/rest/v1",
  "modes": ["data"],
  "vocab": "agrovoc",
  "uri": "http://aims.fao.org/aos/agrovoc/c_8309"
}

Servers you can point it at

• AGROVOC (FAO) — https://agrovoc.fao.org/browse/rest/v1 (multilingual agriculture thesaurus)
• ZBW / STW — https://zbw.eu/beta/skosmos/rest/v1 (economics: STW, JEL classification)
• Loterre — https://loterre.istex.fr/rest/v1 (multi-domain French research vocabularies)
• ACTRIS — https://vocabulary.actris.nilu.no/skosmos/rest/v1 (atmospheric research)
• ...and any other Skosmos instance (they all share the /rest/v1 API).

Output

Each result is one flat row tagged with its mode. Concept rows carry the prefLabel, the full uri, the SKOS type, notation, and (in data mode) the broader / narrower / related URI lists plus every-language labels and mapping links. Results render as a clean table in the Apify Console and export to JSON, CSV, Excel, and more.

Notes

• The data mode keeps all languages of every label (under labels), while prefLabel is the single best label for your chosen lang.
• A search that matches nothing is an honest empty result, not an error. If a whole run produces zero rows across every mode, the actor fails fast rather than reporting false success.
• Only public, read-only SKOS REST endpoints are queried; no authentication is used or required.

Pricing

Pay-per-event: you are charged per result row produced. List vocabularies, search concepts, or resolve a URI — you only pay for the records you get back.

Budget-safe: set a maximum charge for the run and you get exactly the records it pays for — no more. The Actor checks your remaining budget before every Skosmos request, asks search for no more hits than you can pay for, stops as soon as it cannot pay for another record, and never writes unpaid records. If the budget cannot cover a single record, no request is made. If a later mode fails, the records you already paid for are kept.