Triple

T30419813
Position Surface form Disambiguated ID Type / Status
Subject Иоанновский равелин E773866 entity
Predicate охраняется P1040 FINISHED
Object государство Россия 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: государство Россия | Statement: [Иоанновский равелин, охраняется, государство Россия]
PD Predicate disambiguation gpt-5-mini-2025-08-07
Target predicate: охраняется
Context triple: [Иоанновский равелин, охраняется, государство Россия]
  • A. heldToProtect
    Indicates that one entity is kept in someone’s possession or custody for the purpose of safeguarding or protecting it.
  • B. protects chosen
    Indicates taking action to keep someone or something safe from harm, danger, or negative effects.
  • C. guardedIn
    Indicates that one entity is protected or kept under watch within the bounds or premises of another entity.
  • D. keptIn
    Indicates that one entity is stored, contained, or maintained inside another entity as its location or container.
  • E. believedToPreserve
    Indicates that something is thought or assumed to maintain, protect, or keep another thing in its original or desired state.
  • F. None of above.

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_69f22491ba248190b9a4776ca8e42d02 completed April 29, 2026, 3:32 p.m.
NER Named-entity recognition batch_69f6864cd7548190b3e12ef2fe253ad8 completed May 2, 2026, 11:18 p.m.
PD Predicate disambiguation batch_69f678d019fc8190913662cd2f87b857 completed May 2, 2026, 10:21 p.m.
Created at: April 29, 2026, 8:05 p.m.