Triple

T6345536
Position Surface form Disambiguated ID Type / Status
Subject Duntulm Castle E142734 entity
Predicate region P40 FINISHED
Object Highland E4352 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: Highland | Statement: [Duntulm Castle, region, Highland]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Highland
Context triple: [Duntulm Castle, region, Highland]
  • A. Highland
    Highland is a suburban city in Southern California’s Inland Empire, located in San Bernardino County near the city of San Bernardino.
  • B. Highlands region
    The Highlands region is a mountainous inland area of Papua New Guinea known for its rugged terrain, cool climate, and diverse traditional cultures.
  • C. Scottish Highlands chosen
    The Scottish Highlands are a rugged, sparsely populated region in northern Scotland known for their dramatic mountains, deep glens, lochs, and strong Gaelic cultural heritage.
  • D. Kutaisi Lowland
    Kutaisi Lowland is a fertile lowland region in western Georgia known for its agricultural significance and location along the middle course of the Rioni River.
  • E. Grampian
    Grampian was a former local government region in northeast Scotland, centered on the city of Aberdeen and known for its North Sea oil industry and agricultural hinterland.
  • 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_69c008d5ab108190b346c465696824a9 completed March 22, 2026, 3:20 p.m.
NER Named-entity recognition batch_69c067b7e0d08190abacaf41c08ed15e completed March 22, 2026, 10:05 p.m.
NED1 Entity disambiguation (via context triple) batch_69c65fbc66f081909cdf3f33809ee8f1 completed March 27, 2026, 10:45 a.m.
Created at: March 22, 2026, 4:31 p.m.