Triple

T842278
Position Surface form Disambiguated ID Type / Status
Subject Yemen E18202 entity
Predicate hasIsland P970 FINISHED
Object Socotra E58400 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: Socotra | Statement: [Yemen, hasIsland, Socotra]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Socotra
Context triple: [Yemen, hasIsland, Socotra]
  • A. Socotra chosen
    Socotra is a remote Yemeni island renowned for its unique biodiversity and otherworldly landscapes, including the iconic dragon’s blood trees.
  • B. Robert Island
    Robert Island is one of the remote, ice-covered islands of the South Shetland archipelago off the northern tip of the Antarctic Peninsula.
  • C. Madagascar
    Madagascar is a large island nation in the Indian Ocean renowned for its unique biodiversity and high rate of endemic species.
  • D. Qeshm Island
    Qeshm Island is the largest island in the Persian Gulf, known for its strategic location, unique geological formations, and diverse ecosystems off the southern coast of Iran.
  • E. Aldabra Atoll
    Aldabra Atoll is a remote coral atoll in the Indian Ocean’s Seychelles, renowned for its vast population of giant tortoises and largely untouched natural ecosystem.
  • 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_69a4938b04208190b82e1df6b572c548 completed March 1, 2026, 7:29 p.m.
NER Named-entity recognition batch_69a4abe8a0bc81909b54af465e67be1f completed March 1, 2026, 9:13 p.m.
NED1 Entity disambiguation (via context triple) batch_69ac427fffe88190b28bd1b660bb90fe completed March 7, 2026, 3:21 p.m.
Created at: March 1, 2026, 7:38 p.m.