Triple

T18005228
Position Surface form Disambiguated ID Type / Status
Subject Ministerstwo Obrony Narodowej E430728 entity
Predicate headquartersLocation P62 FINISHED
Object Warszawa 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: Warszawa | Statement: [Ministerstwo Obrony Narodowej, headquartersLocation, Warszawa]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Warszawa
Context triple: [Ministerstwo Obrony Narodowej, headquartersLocation, Warszawa]
  • A. Warsaw chosen
    Warsaw is the capital and largest city of Poland, known for its resilient history, especially its near-total destruction in World War II and subsequent postwar reconstruction.
  • B. Warsaw
    Warsaw was the original name of the English post-punk band that later became known as Joy Division.
  • C. Warsaw
    Warsaw is a small city in northern Indiana known as the "Orthopedic Capital of the World" for its concentration of medical device manufacturers.
  • D. Lublin
    Lublin is a historic city in eastern Poland known as a major cultural, academic, and economic center and for its significant role in Polish political history.
  • E. Ville de Varsovie
    Ville de Varsovie was a French ship of the line that took part in the early 19th-century naval engagement known as the Battle of the Basque Roads.
  • 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_69d8b904530081908bf341d842464856 completed April 10, 2026, 8:47 a.m.
NER Named-entity recognition batch_69e4b51ba1888190a339d726e92f376b completed April 19, 2026, 10:57 a.m.
Created at: April 10, 2026, 10:24 a.m.