Triple

T5011480
Position Surface form Disambiguated ID Type / Status
Subject Gabrielle Starr E112630 entity
Predicate previouslyHeldPosition P18982 FINISHED
Object dean of the College of Arts and Science at New York University LITERAL 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: dean of the College of Arts and Science at New York University | Statement: [Gabrielle Starr, previouslyHeldPosition, dean of the College of Arts and Science at New York University]
PD Predicate disambiguation gpt-5-mini-2025-08-07
Target predicate: previouslyHeldPosition
Context triple: [Gabrielle Starr, previouslyHeldPosition, dean of the College of Arts and Science at New York University]
  • A. officePreviouslyHeldBy chosen
    Indicates that a particular office or position was formerly occupied by a specified person or entity.
  • B. positionHeld
    Indicates that an entity occupies or has occupied a specific role, job, office, or position within an organization or context.
  • C. hasPreviouslyBeenHeldIn
    Indicates that an entity was located, confined, or kept in a particular place or container at some time in the past.
  • D. mayHavePriorRole
    Indicates that an entity is allowed or expected to have held a specified role at some earlier time.
  • E. heldJudicialPositionIn
    Indicates that an entity served in an official judicial role or office within a specified jurisdiction or court.
  • 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_69bd4434acb8819086679dbeccc2fe54 completed March 20, 2026, 12:57 p.m.
NER Named-entity recognition batch_69bd730d70ac8190a1818d628e95ab04 completed March 20, 2026, 4:17 p.m.
PD Predicate disambiguation batch_69bd714cbc448190aa53a8a83d768b64 completed March 20, 2026, 4:09 p.m.
Created at: March 20, 2026, 1:35 p.m.