Triple

T25342007
Position Surface form Disambiguated ID Type / Status
Subject Korean Air Lines Flight 007 shootdown E635436 entity
Predicate interceptingPilot P133688 FINISHED
Object Gennadi Osipovich 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: Gennadi Osipovich | Statement: [Korean Air Lines Flight 007 shootdown, interceptingPilot, Gennadi Osipovich]
PD Predicate disambiguation gpt-5-mini-2025-08-07
Target predicate: interceptingPilot
Context triple: [Korean Air Lines Flight 007 shootdown, interceptingPilot, Gennadi Osipovich]
  • A. pilotHijacker
    Indicates that the subject is a pilot who hijacks, or has hijacked, an aircraft.
  • B. interceptedBy
    Indicates that an action, communication, or movement is stopped, captured, or diverted by another agent before reaching its intended target or destination.
  • C. pilotFor chosen
    Indicates that one entity serves as the pilot operating or navigating a vehicle, aircraft, or vessel on behalf of another entity (such as an organization, mission, or owner).
  • D. interceptingPlayer
    Indicates that one player takes possession of or disrupts a pass, throw, or transmission intended for another player or target.
  • E. interceptRepresents
    Indicates that one entity acts as a representation or model of the interception of another entity or process.
  • F. None of above.

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_69e75a99bd6481909476115b35b9a8e4 completed April 21, 2026, 11:08 a.m.
NER Named-entity recognition batch_69f621fcea1481909b6f8b3af1ee6820 completed May 2, 2026, 4:10 p.m.
PD Predicate disambiguation batch_69f620dc38088190b56b2b15ed75b3c2 completed May 2, 2026, 4:05 p.m.
Created at: April 21, 2026, 1:32 p.m.