Triple

T23264875
Position Surface form Disambiguated ID Type / Status
Subject Pat Roach E588117 entity
Predicate name P16 FINISHED
Object Pat Roach NE NERFINISHED

How this triple was built (2 steps)

Every LLM step that produced this triple, in pipeline order — named-entity classification, the disambiguation choices (the exact options shown, with the pick highlighted), and the generated description. The batch + timestamp of each is in the Provenance table below.

NER Named-entity recognition gpt-5-mini
Instruction
Given a phrase, classify it is english named entity (e.g., persons, organizations, works of art) in Latin script, or not (e.g., literals, dates, URLs, verbose phrases). For disambiguation, the statement where the phrase occurs as object is also given. Please return a JSON object with `phrase` (string, the phrase being analyzed) and `is_ne` (boolean, indicating whether the phrase is a Named Entity).
Input
Phrase: Pat Roach | Statement: [Pat Roach, name, Pat Roach]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Pat Roach
Context triple: [Pat Roach, name, Pat Roach]
  • A. Pat Roach chosen
    Pat Roach was a British professional wrestler and character actor best known for his physically imposing roles in films like the Indiana Jones series and the TV show "Auf Wiedersehen, Pet."
  • B. Patrick Roach
    Patrick Roach is a Canadian actor best known for playing the perpetually shirtless, alcoholic character Randy on the mockumentary TV series Trailer Park Boys.
  • C. Steve Reaves
    Steve Reaves is a producer known for his work on the film "The Keynote Speaker."
  • D. Chris Doleman
    Chris Doleman was a Hall of Fame American football defensive end renowned for his prolific pass rushing during a long NFL career, primarily with the Minnesota Vikings.
  • E. Brad Van Pelt
    Brad Van Pelt was a standout linebacker for the New York Giants in the 1970s and early 1980s, known for his Pro Bowl-caliber play and leadership on defense.
  • F. None of above.
  • G. Unsure - the case is ambiguous/there is not enough information to decide.

Provenance (2 batches)

The batch behind each pipeline step, in order, with when it ran. Timestamps are batch-level — stages were processed in waves, so the object chain (NER → NED1 → NEDg → NED2) reads in order, but predicate / elicitation batches can sit in a different wave.

Step Stage Batch ID Status When
creating Elicitation batch_69e25d148adc819088efbf42672604e9 completed April 17, 2026, 4:17 p.m.
NER Named-entity recognition batch_69f194cb76c48190869915cd93b44fcc completed April 29, 2026, 5:19 a.m.
Created at: April 17, 2026, 4:29 p.m.