Triple

T23514493
Position Surface form Disambiguated ID Type / Status
Subject Mangareva E574317 entity
Predicate hasNearbyIsland P970 FINISHED
Object Taravai 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: Taravai | Statement: [Mangareva, hasNearbyIsland, Taravai]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Taravai
Context triple: [Mangareva, hasNearbyIsland, Taravai]
  • A. Taravai chosen
    Taravai is one of the inhabited islands of the Gambier Islands archipelago in French Polynesia, known for its remote location and traditional Polynesian character.
  • B. Tashir
    Tashir is a small town located in Armenia’s northern Lori Province.
  • C. Tarikere
    Tarikere is a town in the Chikkamagaluru district of Karnataka, India, known as a regional hub for agriculture and as a gateway to nearby hills, forests, and reservoirs.
  • D. Tarare
    Tarare is a commune in eastern France known historically for its textile industry and its location in the hilly Beaujolais region of the Rhône department.
  • E. Taraval
    Taraval is a major east–west street in San Francisco’s Sunset District, known for its residential character and Muni light rail service.
  • 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_69e245bb3dcc8190ba9a2b35972b58d0 completed April 17, 2026, 2:37 p.m.
NER Named-entity recognition batch_69f1aa80d9048190ab735dddd301feb4 completed April 29, 2026, 6:51 a.m.
Created at: April 17, 2026, 6:08 p.m.