Linked data

This site is more than pages — it is a published knowledge graph: entities and their relationships described as RDF triples, offered in several downloadable, queryable serializations. This page is built around Turtle (/index.ttl), the most readable form; JSON-LD and N-Triples are the same triples in another syntax.

Serializations & access points

One graph, the same triples, in forms suited to each use:

The three serializations are isomorphic — exactly the same triples, different syntax. Content negotiation by Accept header also works on any page URL. Every machine-discovery resource is listed in the discovery catalog.

Dataset DOI & citation

This site's public knowledge graph is registered with Crossref as a citable dataset.

RDF 1.2 & SPARQL 1.2

Two RDF 1.2 claims come from the home graph. rdf:reifies links each to an exact authored passage, a SHA-256 addressed source version, publisher, issue date, status, and the change log. The build verifies both the graph claims and source passages, then runs the query with pinned PurRDF.

• RDF 1.2 Turtle claim graph
• SPARQL 1.2 query · computed result
• Manifest and canonical identity

purrdf query --data claims.ttl "$(cat claims.rq)"

Run the query and view the graph in this browser

The WebAssembly engine loads on demand; the query runs in this browser.

Query the full site graph

Run a SPARQL query in this browser over the published /index.ttl. The engine loads only when you submit a query.

The claim example is a separate RDF 1.2 dataset. The main graph's /sparql endpoint uses PurRDF; remote SERVICE federation is unavailable until the async resolver bridge ships.

What it's good for

• Query it with SPARQL instead of scraping pages.
• Reconcile entities — grounded with Wikidata QIDs (and ORCID, DBLP…) so they align with external datasets.
• Reuse the whole dataset — licensed Creative Commons BY-NC-SA CAv2.5, downloadable and redistributable.
• Validate it against SHACL and ShEx shapes (below).
• Feed agents — the same graph is exposed to AI clients via MCP and llms.txt.

Key sections of the graph

Every entity is identified by an IRI fragment (#…) in the root Turtle document. The root person node is #paudley.

• Expertise taxonomy (SKOS): 33 skos:Concept nodes under the #kw-scheme scheme, labelled in en/fr/zh, Wikidata-grounded and aligned to ACM, IEEE SWEBOK, LCSH, ESCO, EuroVoc and more.
• Publications & patents: 3 publications and 2 patents (BIBO/BIBFRAME, with DOIs and sameAs).
• Open-source projects: described with DOAP (repository, licence, language).
• Genealogy dataset: a family graph in the GEDCOM vocabulary, dedicated to the public domain (CC0); living individuals reduced to name and kinship.
• Social & media: posts and podcasts as SIOC and schema CreativeWork, plus media coverage.

What one SKOS concept looks like in Turtle:

@prefix skos: <http://www.w3.org/2004/02/skos/core#> .
@prefix owl:  <http://www.w3.org/2002/07/owl#> .
@prefix wd:   <http://www.wikidata.org/entity/> .
@prefix lcsh: <http://id.loc.gov/authorities/subjects/> .

<#concept-semantic-web> a skos:Concept ;
    skos:inScheme    <#kw-scheme> ;
    skos:prefLabel   "Semantic Web"@en, "Web sémantique"@fr, "语义网"@zh ;
    owl:sameAs       wd:Q54837 ;
    skos:closeMatch  lcsh:sh2002000569 .

Ontologies overview

The graph reuses established standard vocabularies, grouped here by purpose:

Shapes, validation & self-description

The graph is constrained and self-describing — its structure is declared in two standard shape languages, with dataset metadata:

• /.well-known/shacl.ttl — SHACL constraint shapes
• /.well-known/shex.shex — ShEx shape expressions
• /.well-known/void.ttl — VoID dataset description

Example query — every concept with its Wikidata grounding:

PREFIX skos: <http://www.w3.org/2004/02/skos/core#>
PREFIX owl:  <http://www.w3.org/2002/07/owl#>
SELECT ?concept ?label ?wikidata WHERE {
  ?concept a skos:Concept ; skos:prefLabel ?label .
  OPTIONAL { ?concept owl:sameAs ?wikidata }
}