Triple

T20088221
Position Surface form Disambiguated ID Type / Status
Subject George Allen E496192 entity
Predicate father P120 FINISHED
Object George Herbert Allen 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: George Herbert Allen | Statement: [George Allen, father, George Herbert Allen]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: George Herbert Allen
Context triple: [George Allen, father, George Herbert Allen]
  • A. George Herbert Allen chosen
    George Herbert Allen was the father of American football coach and NFL Hall of Famer George Allen.
  • B. George Herbert Allen
    George Herbert Allen is an American politician and attorney who served as Governor of Virginia and later as a United States Senator.
  • C. Carney Lansford
    Carney Lansford is a former Major League Baseball third baseman best known for his All-Star career with the Boston Red Sox and Oakland Athletics, including winning a World Series with the A's in 1989.
  • D. Tom Fitton
    Tom Fitton is an American conservative activist and commentator best known as the longtime president of the watchdog organization Judicial Watch.
  • E. Howard A. Smith
    Howard A. Smith was a film editor best known for his work on classic Hollywood movies, including the 1961 romantic comedy "Breakfast at Tiffany's."
  • 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_69da626eee3881909f3454986d4a6511 completed April 11, 2026, 3:02 p.m.
NER Named-entity recognition batch_69e6655c7de081909a400c736e92495d completed April 20, 2026, 5:41 p.m.
Created at: April 11, 2026, 11:16 p.m.