Triple
T24952045
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | 北小島 |
E624365
|
entity |
| Predicate | 最寄りの有人島 |
P72701
|
FINISHED |
| Object | 青ヶ島 |
—
|
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: 青ヶ島 | Statement: [北小島, 最寄りの有人島, 青ヶ島]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: 最寄りの有人島 Context triple: [北小島, 最寄りの有人島, 青ヶ島]
-
A.
nearestPopulatedIsland
chosen
Indicates that one island is the closest populated island to another specified location or island.
-
B.
nearbyAtoll
Indicates that one entity (typically a location or feature) is situated close to or in the vicinity of an atoll.
-
C.
hasNearbyMainIsland
Indicates that one entity has a main island located in close geographic proximity to it.
-
D.
onlyInhabitedIsland
Indicates that the subject is the sole inhabited island within a specified geographic or contextual set, with no other islands in that set having inhabitants.
-
E.
associatedIsland
Indicates that one entity is related to or connected with a particular island in some relevant way.
- 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_69e2ff22e4c48190a0444b5a044f14e8 |
completed | April 18, 2026, 3:48 a.m. |
| NER | Named-entity recognition | batch_69f424004d588190abe0115931aab67a |
completed | May 1, 2026, 3:54 a.m. |
| PD | Predicate disambiguation | batch_69f4210130d08190ae30b7943f7a0bbc |
completed | May 1, 2026, 3:41 a.m. |
Created at: April 18, 2026, 5:57 a.m.