Dogfooding was one of RDF’s biggest challenges prior to the arrival of LLMs as powerful general-purpose clients. Why? Because transforming and presenting RDF specifications in RDF form was difficult. Today, that problem is gone. Here’s an example of the new RDF 1.2 primer, deployed as a knowledge graph that uses Linked Data principles to manifest a Semantic Web.
Why Should You Care About Dogfooding RDF 1.2 Specs?
You can ask a variety of questions via any AI agent that supports skills against the generated knowledge graph. All you need to do is load the Data Twingler skills and query using natural language, without being tripped up or distracted by hallucinations (since these skills include a KG-only mode that ensures no access to an LLM’s native graph unless you explicitly enable that modality).
Live Demo Links:
[1] https://linkeddata.uriburner.com/DAV/demos//daas/rdf12-primer-deepseek_v4pro-1.html -- HTML document that breakdown the RDF 1.2 primer doc
[2] https://linkeddata.uriburner.com/DAV/demos//daas/rdf12-primer-deepseek_v4pro-1.md -- Markdown rendition of HTML doc
[3] https://linkeddata.uriburner.com/describe/?url=https%3A%2F%2Fwww.w3.org%2FTR%2Frdf12-primer%2F%23article -- Faceted Search & Browsing based HTML page
[4] https://linkeddata.uriburner.com/c/9u2mBAKo6Y -- SPARQL Query Results Page
In all cases, just click to explore using the follow-your-nose pattern
Tools used:
1. Anthropic’s Claude Code as the AI agent
2. DeepSeek-v4 as the LLM
3. Virtuoso Data Spaces platform, comprising:
* A WebDAV-based filesystem for hosting generated RDF, HTML, and Markdown documents
* A high-performance DBMS engine that automatically ingests and manages these documents
* Support for RDF graphs and relational tables accessible via SQL, SPARQL, GraphQL, etc. (note: we also have openCypher and GQL in the works to bury all distractions)
Kingsley Uyi Idehen