Triple

T17331968
Position Surface form Disambiguated ID Type / Status
Subject Death Wish E420837 entity
Predicate director P255 FINISHED
Object Michael Winner E167698 NE FINISHED

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: Michael Winner | Statement: [Death Wish, director, Michael Winner]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Michael Winner
Context triple: [Death Wish, director, Michael Winner]
  • A. Michael Winner chosen
    Michael Winner was a British film director and producer best known for directing the "Death Wish" series starring Charles Bronson.
  • B. Gered Mankowitz
    Gered Mankowitz is a British photographer best known for his iconic music and rock-and-roll imagery, including portraits and album covers for major artists.
  • C. Lew Grade
    Lew Grade was a prominent British media mogul and television producer who played a key role in shaping commercial television in the UK during the mid-20th century.
  • D. Ronald Neame
    Ronald Neame was a British film director, producer, and cinematographer known for works such as "The Poseidon Adventure" and collaborations with Alec Guinness.
  • E. Ian McKay
    Ian McKay was a British Army sergeant posthumously awarded the Victoria Cross for exceptional bravery during the Falklands War.
  • F. None of above.
  • G. Unsure - the case is ambiguous/there is not enough information to decide.

Provenance (3 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_69d889d3adc881909319f1edb8d2a956 completed April 10, 2026, 5:25 a.m.
NER Named-entity recognition batch_69e439d6870c8190989897aa6beba8ff completed April 19, 2026, 2:11 a.m.
NED1 Entity disambiguation (via context triple) batch_6a018c5025d08190ab2581a3b04ae661 completed May 11, 2026, 7:59 a.m.
Created at: April 10, 2026, 5:43 a.m.