Graphs

In a SPARQL dataset, every fact (triple) lives in a graph — and the same fact can live in several named graphs at once. That "where does this fact come from" is real information, so AE RDF always knows the graph(s) behind every triple it shows you, even when it isn't painting them on screen.

Where you see it

• Resource header — a summary of the named graph(s) the resource's facts come from (or "default graph").
• Show graphs toggle — next to the summary. Off by default to keep the view clean; turn it on to see the graph on every triple.
• Multi-graph facts are always flagged — if a single triple is asserted in more than one graph, it's badged even when the toggle is off, because hiding that would be misleading.
• Hover — hovering any value shows its graph(s) regardless of the toggle.

How it works

AE RDF understands an endpoint on two simple axes, and builds every query from them:

• Quads? — does it use named graphs at all? Resolved automatically (reused from AE SKOS if known, else a quick probe on connect), or set in config.
• The explicit (default) view — is it the endpoint's own triples, or just a merged view of the quads? When it's merely merged (common with Virtuoso, e.g. CORDIS), that view is redundant and returns each fact once per graph — so AE RDF never queries it, reading the named graphs directly and de-duplicating. That's why values aren't doubled.

The second axis isn't something a tool can reliably guess, so it's part of the endpoint config. With Config authoring mode on, edit an endpoint in the Endpoint Manager and use its Graph behaviour section to set Named graphs (quads) and Default view (Auto / Own / Merged); it's saved with the endpoint and included in the exported app.json. Declare it once per deployment and every query is correct and fast. Unset, AE RDF plays it safe (queries everything and de-duplicates).

Not yet included

• Picking/filtering by graph (show only one graph, hide others) — AE RDF shows provenance, it doesn't yet let you scope to a graph.