Triple

T1669423
Position Surface form Disambiguated ID Type / Status
Subject Savin Hill E36089 entity
Predicate locatedIn P40 FINISHED
Object Dorchester E73521 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: Dorchester | Statement: [Savin Hill, locatedIn, Dorchester]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Dorchester
Context triple: [Savin Hill, locatedIn, Dorchester]
  • A. Dorchester chosen
    Dorchester is one of the earliest English settlements in New England, now a historic neighborhood of Boston, Massachusetts.
  • B. Roxbury
    Roxbury is one of the earliest English-settled neighborhoods of present-day Boston, historically significant as a colonial town in what became Massachusetts.
  • C. Charlestown
    Charlestown is a district in north Manchester, England, known as a primarily residential area with local amenities and community facilities.
  • D. Charlestown
    Charlestown is the main town and administrative center of the Caribbean island of Nevis, known for its colonial architecture and historic hot springs.
  • E. Jamaica Plain
    Jamaica Plain is a residential neighborhood in Boston, Massachusetts, known for its diverse community, green spaces, and historic architecture.
  • 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_69a8861286808190939afff3ce8ee31e completed March 4, 2026, 7:20 p.m.
NER Named-entity recognition batch_69a90adf3d3c81909233e574e79b82a2 completed March 5, 2026, 4:47 a.m.
NED1 Entity disambiguation (via context triple) batch_69b0fc189de08190b88849333d824c5b completed March 11, 2026, 5:22 a.m.
Created at: March 4, 2026, 7:29 p.m.