Triple

T3850505
Position Surface form Disambiguated ID Type / Status
Subject Beria period of the NKVD E85282 entity
Predicate terminatesWith P3587 FINISHED
Object reorganization of NKVD into MVD and MGB 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: reorganization of NKVD into MVD and MGB | Statement: [Beria period of the NKVD, terminatesWith, reorganization of NKVD into MVD and MGB]
PD Predicate disambiguation gpt-5-mini-2025-08-07
Target predicate: terminatesWith
Context triple: [Beria period of the NKVD, terminatesWith, reorganization of NKVD into MVD and MGB]
  • A. blendsWith
    Indicates that one entity can be mixed or combined smoothly with another to form a uniform or harmonious result.
  • B. endsWith
    Indicates that one entity terminates with, or has as its final part, the sequence or element represented by the other entity.
  • C. hasEnding
    Indicates that one entity concludes with, or terminates in, another entity (such as a specific substring, segment, or final component).
  • D. hasTerminusIn
    Indicates that something (such as a route, line, or path) ends or has its final stopping point at a specified location.
  • E. endedWith chosen
    Indicates that one event, process, or state concluded with or was finalized by another specified event, condition, or outcome.
  • 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_69aed936de1c81908f91bed80f70abb2 completed March 9, 2026, 2:29 p.m.
NER Named-entity recognition batch_69aeebcf67788190975105131baabc4b completed March 9, 2026, 3:48 p.m.
PD Predicate disambiguation batch_69aee750377c8190af70c79768c0edd8 completed March 9, 2026, 3:29 p.m.
Created at: March 9, 2026, 3:19 p.m.