Intentional Arrangement SKOS

Build SKOS taxonomies and thesauri on open standards. Glossary, editor, business view, qSKOS validation, visualizer, DCMI metadata, SPARQL and import/export, with provenance and governance.

Untitled ✓ Saved
Concept scheme, identifiers & Dublin Core metadata
Concept identifiers
Label style
ISO 25964 relations
Attribution — link an agent
Point creator/contributors/publisher at agents defined in the Sources tab (prov:Agent). Contributors repeat — one dcterms:contributor triple per agent, per DCMI practice. An agent reference overrides the plain-text field on export.

Concepts

Concept editor

Select a concept on the left, or add a top concept to begin.

Proposed terms

Collections

Group concepts as skos:Collection / skos:OrderedCollection — alternative groupings that sit alongside the broader/narrower hierarchy, not inside it.

Collection editor

Sources

Reusable documents (foaf:Document) and agents (prov:Agent). Link a concept to a document with dcterms:source in the concept editor; credit an agent in the scheme metadata.

Source editor

Select a document or agent on the left, or add one above.
⇄
Export
v0.18.13 · Align concepts across two taxonomies with SKOS mapping properties to build a federated thesaurus. Auto-match proposes links from identical/near labels; confirm or add your own, then export the crosswalk or the full thesaurus. Mappings are stored on the source concepts, so they also flow into normal RDF export. Mapping to an external vocabulary (SKOS itself, Schema.org, Dublin Core…)? Add it on the concept — Build tab → External mappings — where you can type a full IRI or a term like skos:Concept.
Align · map · federate

Crosswalk taxonomies into a thesaurus

Pick a second taxonomy (another project in your workspace) and align its concepts with this one using the five SKOS mapping relations — exactMatch, closeMatch, broadMatch, narrowMatch, relatedMatch. Auto-match seeds the alignment from matching labels; you confirm and refine. Export the crosswalk on its own, or the whole federated thesaurus (both schemes + the mappings) as one interoperable RDF file.

A crosswalk needs two taxonomies. Only have one? Import another, or

Nothing to show yet. Add a Concept or Class, or build your taxonomy/ontology in the other tabs — it appears here automatically.
Drag bubbles · scroll to zoom · drag canvas to pan · Connect then click two bubbles to link them
concept class vocabulary scheme source agent mapped
Federation — all crosswalked taxonomies
One file with every taxonomy in your workspace that participates in a crosswalk — all schemes, all concepts, all SKOS mapping triples.
Share a read-only link
Copy a link that opens this taxonomy in a read-only Business view — the taxonomy travels inside the link, so there's no server or account. Anyone with the link can browse it.
Publish to a server (Apache Jena Fuseki)
Store this taxonomy in a triplestore and get a short link that loads it from the server — no size limit, SPARQL-queryable. Needs a Fuseki endpoint (connect one).
Format
Labeling
Metadata 

    

  


  
  

    

      

        Mode
        

        
        
        

        
      

      
      
Run validation to check your taxonomy against qSKOS quality criteria.

      

    

  


  
  
  

    

      

        
Glossary

        
        
        
        
        
        
      

      

        
        

          
          
        

      

      

      

        
        
        
      

    

    
A staging bucket for unstructured candidate terms — a garage before the taxonomy. Import a flat list, link it to a taxonomy, then promote terms into it as SKOS concepts — each one as a top concept or placed under a parent you choose. Lists can also sit unlinked until you're ready.

  


  
  
  

    

      

        Ask in plain language
        
        
      

      

        
        
        ⌘/Ctrl + Enter to run · runs locally in your browser
        
      

      
      

        

        

      

    

    
The natural-language box maps common questions to SPARQL you can edit. Supported: SELECT/DISTINCT, BGP, a, prefixes & IRIs, OPTIONAL, FILTER, single-predicate paths p+/p*, COUNT/SUM with GROUP BY, ORDER BY, LIMIT. Prefix : is your base namespace. Everything runs in your browser — nothing is uploaded.