Triple
T24563218
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Hamburg–Westerland railway |
E607711
|
entity |
| Predicate | hasEndpointOnIsland |
P133195
|
FINISHED |
| Object | Sylt |
—
|
NE NERFINISHED |
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: Sylt | Statement: [Hamburg–Westerland railway, hasEndpointOnIsland, Sylt]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: hasEndpointOnIsland Context triple: [Hamburg–Westerland railway, hasEndpointOnIsland, Sylt]
-
A.
isOnRemoteIsland
Indicates that an entity is located on an island that is geographically isolated or difficult to access from mainlands or populated areas.
-
B.
hasIslandConnection
chosen
Indicates that there exists a connection or link between entities that specifically involves or passes through an island.
-
C.
hasIsland
Indicates that one entity possesses, contains, or includes an island as part of its domain, territory, or structure.
-
D.
isInIsland
Indicates that one entity is located within the geographic boundaries of an island.
-
E.
isOnSideOfIsland
Indicates that one entity is located on a particular side or shore of an island relative to some reference or division.
- 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_69e2c4cc35a48190990b7571bc086df8 |
completed | April 17, 2026, 11:39 p.m. |
| NER | Named-entity recognition | batch_69f2a8f8a7488190a11f32014539a4a3 |
completed | April 30, 2026, 12:57 a.m. |
| PD | Predicate disambiguation | batch_69f2a6b99e7c8190ba7e2dc8729a314a |
completed | April 30, 2026, 12:47 a.m. |
Created at: April 18, 2026, 2:28 a.m.