eyeprolog

RDF 1.2 graph join

Which station reported this sensor reading?

Sensor readings sit in a named graph. A triple in the default graph says which station reported that whole graph. One rule joins the two levels, so each sensor is linked to its reporting station.

Run it in the playground · Source

---

The question

Which station is responsible for each sensor’s data?

This is one of a family of RDF 1.2 examples. RDF data arrives as rdf(Subject, Predicate, Object, Graph) facts, as produced by rdf-prolog-interchange. In TriG, a graph name can itself be the subject of triples, which is how datasets record provenance about whole graphs.

---

The data

rdf(iri('https://example.org/sensorGraph'), iri('https://example.org/reportedBy'), iri('https://example.org/station7'), default_graph).
rdf(iri('https://example.org/sensor1'), iri('https://example.org/temperature'), literal('21.5', datatype('http://www.w3.org/2001/XMLSchema#string')), iri('https://example.org/sensorGraph')).

The first triple is metadata about a graph: sensorGraph was reported by station7. The second is data inside that graph.

---

The rule

result_rdf(S, iri('https://example.org/reportedBy'), Station, default_graph) :-
  rdf(
    iri('https://example.org/sensorGraph'),
    iri('https://example.org/reportedBy'),
    Station,
    default_graph
  ),
  rdf(
    S,
    iri('https://example.org/temperature'),
    _,
    iri('https://example.org/sensorGraph')
  ).

The first goal reads the graph’s metadata; the second finds the sensors whose readings are in that graph. The join goes through the graph name.

---

The answer

eyeprolog examples/rdf12-trig-graph-join.pl

result_rdf(iri('https://example.org/sensor1'), iri('https://example.org/reportedBy'), iri('https://example.org/station7'), default_graph).

---

The reason

step(result_rdf(iri('https://example.org/sensor1'), iri('https://example.org/reportedBy'), iri('https://example.org/station7'), default_graph),
     rule(3),
     ['S' = iri('https://example.org/sensor1'), 'Station' = iri('https://example.org/station7')],
     [rdf(iri('https://example.org/sensorGraph'), iri('https://example.org/reportedBy'), iri('https://example.org/station7'), default_graph),
      rdf(iri('https://example.org/sensor1'), iri('https://example.org/temperature'), literal('21.5', datatype('http://www.w3.org/2001/XMLSchema#string')), iri('https://example.org/sensorGraph'))]).

The conclusion uses two premises from two different graphs, and the proof names both.

---

The check

condition('C1', resolution, ok, 3).
condition('C4', coverage, ok, 3).
condition('C7', relevance, ok, 4).
trusted(0).
verdict(checked).

---

Try it

• Replace the fixed sensorGraph IRI in the rule with a variable, so every reported graph is joined with its own readings.
• Add a second sensor to sensorGraph and predict the output.
• Compare with named-graph reasoning, where rules stay inside one graph.

Learn more in The Art of EyeProlog, Chapter 15.

This site is open source. Improve this page.