Triple

T4904172
Position Surface form Disambiguated ID Type / Status
Subject Tbilisi Metro E109873 entity
Predicate connectsDistrict P2564 FINISHED
Object Didube E478956 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: Didube | Statement: [Tbilisi Metro, connectsDistrict, Didube]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Didube
Context triple: [Tbilisi Metro, connectsDistrict, Didube]
  • A. Didube chosen
    Didube is a metro station in Tbilisi, Georgia, serving as a key transit hub that connects the city’s metro system with major bus and marshrutka routes to other regions.
  • B. Jebba
    Jebba is a town in western Nigeria known for its strategic location on the Niger River and its historic bridge linking northern and southern Nigeria.
  • C. Dschubba
    Dschubba is a bright multiple star system that forms the "head" of the constellation Scorpius and is designated Delta Scorpii.
  • D. Kibibi
    Kibibi is a lively and energetic host character in Disney's "Festival of the Lion King" stage show at Disney theme parks.
  • E. Hadibu
    Hadibu is the main town and administrative center of the Yemeni island of Socotra in the Arabian Sea.
  • 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_69bd441180708190ba42ffb44fea533a completed March 20, 2026, 12:56 p.m.
NER Named-entity recognition batch_69bd6e6fdeac81909092f51ae40ad20e completed March 20, 2026, 3:57 p.m.
NED1 Entity disambiguation (via context triple) batch_69be779c47348190bd8e19f87c2aa4c2 completed March 21, 2026, 10:49 a.m.
Created at: March 20, 2026, 1:29 p.m.