Triple

T23366522
Position Surface form Disambiguated ID Type / Status
Subject Raiatea E593337 entity
Predicate rankBySizeInFrenchPolynesia P152444 FINISHED
Object second largest island after Tahiti 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: second largest island after Tahiti | Statement: [Raiatea, rankBySizeInFrenchPolynesia, second largest island after Tahiti]
PD Predicate disambiguation gpt-5-mini-2025-08-07
Target predicate: rankBySizeInFrenchPolynesia
Context triple: [Raiatea, rankBySizeInFrenchPolynesia, second largest island after Tahiti]
  • A. rankByAreaInSeychelles
    Indicates the relative ordering of entities based on their area size specifically within the context of Seychelles.
  • B. rankInSizeInNewZealandIslands
    Indicates the ordinal position of an island in New Zealand when ordered by size (e.g., largest, second largest, etc.).
  • C. rankInArchipelagoByArea
    Indicates the position of an island in a size-based ordering of islands within the same archipelago, based on their land area.
  • D. rankingInCountryBySize
    Indicates the position of an entity in an ordered list of entities within a specific country, based on their relative size.
  • E. rankingBySizeUSIslands
    Indicates the relative order of U.S. islands based on their size, from largest to smallest (or vice versa).
  • F. None of above. chosen

Provenance (4 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_69e25d2593c88190bcdf4a716a94ccb2 completed April 17, 2026, 4:17 p.m.
NER Named-entity recognition batch_69f1a0ac7494819082e98ac5632eba28 completed April 29, 2026, 6:09 a.m.
PD Predicate disambiguation batch_69f061c7aaa48190a58ce93f87155ffc completed April 28, 2026, 7:29 a.m.
PDg Predicate description generation batch_69f0bd4a0e408190ad8916faf23562d9 completed April 28, 2026, 1:59 p.m.
Created at: April 17, 2026, 5:31 p.m.