Annotated claims

Two sources disagree. Which one do we follow, and why?

rdf12-annotated-claims.js · output · proof · check · try it in the playground

The question

The transport authority says the bridge is closed, with confidence 98. An anonymous post says it is open, with confidence 35. The authority’s trust score is 95, the post’s 20. Each claim is an RDF 1.2 reifier of its status triple.

Should we cross the bridge?

What we tell eyecode

• A claim is an asserted status, a reifier of it, its source, and a score: confidence times trust.
• A claim’s rank is one more than the number of claims that score higher.
• The top claim decides, but only when its score reaches 8000; otherwise the bridge must be checked by hand.

Numbers in RDF are typed literals whose value is text. number_chars reads the text as a number, and spells a computed number back as text:

implied_by(integer_value(literal(Text, datatype(integer)), N),
           [atom_chars(Text, Chars), number_chars(N, Chars)])

RDF values are ordinary terms: iri(Text) for an IRI, literal(Text, datatype(IRI)) or literal(Text, lang(Tag)) for a literal, triple(S, P, O) for a triple term, and rdf(S, P, O, Graph) for a quad in a graph, 'default_graph' or iri(Name).

What eyecode concludes

The decision, and the ranked evidence behind it:

result_rdf(iri('https://example.org/bridge'), iri('https://example.org/decision'), iri('https://example.org/avoid_bridge'), 'default_graph')
result_rdf(iri('https://example.org/bridge'), iri('https://example.org/trustedSource'), iri('https://example.org/transportAuthority'), 'default_graph')
result_rdf(iri('https://example.org/bridge'), iri('https://example.org/score'), literal('9310', datatype('http://www.w3.org/2001/XMLSchema#integer')), 'default_graph')
result_rdf(iri('https://example.org/closed'), iri('https://example.org/rank'), literal('1', datatype('http://www.w3.org/2001/XMLSchema#integer')), 'default_graph')
result_rdf(iri('https://example.org/closed'), iri('https://example.org/assertedBy'), iri('https://example.org/transportAuthority'), 'default_graph')
result_rdf(iri('https://example.org/closed'), iri('https://example.org/score'), literal('9310', datatype('http://www.w3.org/2001/XMLSchema#integer')), 'default_graph')
result_rdf(iri('https://example.org/open'), iri('https://example.org/rank'), literal('2', datatype('http://www.w3.org/2001/XMLSchema#integer')), 'default_graph')
result_rdf(iri('https://example.org/open'), iri('https://example.org/assertedBy'), iri('https://example.org/anonymousPost'), 'default_graph')
result_rdf(iri('https://example.org/open'), iri('https://example.org/score'), literal('700', datatype('http://www.w3.org/2001/XMLSchema#integer')), 'default_graph')

Why: the proof in plain words

1. The authority’s claim scores 98 × 95 = 9310; the post’s 35 × 20 = 700 — the claim rule, with the multiplications recomputed.
2. No claim scores higher than 9310, so closed ranks first; one does for 700, so open ranks second — the rank rule.
3. 9310 reaches 8000, and closed means avoid the bridge — the decision rule.

Each step names the exact line of the program it used, and the values it filled in.

Checked, with obligations

• 60 steps for 9 results: 36 verified against the program lines they cite, and 22 calculations, comparisons and number readings recomputed.
• 2 steps are taken on trust: the two collected obligations, the lists of higher scores used for each rank, [] for 9310 and [9310] for
  1. That each list is complete is recorded, not proved.
• The checker tries to refute both against the evidence and finds no contradiction.

Verdict: checked_with_obligations.

Try it

eyecode examples/rdf12-annotated-claims.js                  # the results as quads
eyecode --proof examples/rdf12-annotated-claims.js          # with their proof
eyecode examples/rdf12-annotated-claims.js > out.js && eyecode-to-rdf out.js -o out.ttl

eyecode-to-rdf and rdf-to-eyecode come with rdf-interchange; they turn eyecode’s quads into RDF and RDF into quads. Or open it in the playground. Lower the authority’s confidence to 80 and run again: 7600 is below the threshold, and the decision becomes verify_manually.

Takeaway

When sources conflict, the reasons for following one of them can be written down, computed, ranked and checked, not just asserted.