Triple
T10729803
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Eagle Tower |
E253041
|
entity |
| Predicate | relativeSizeWithinCastle |
P94423
|
FINISHED |
| Object | one of the largest towers of Caernarfon Castle |
—
|
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: one of the largest towers of Caernarfon Castle | Statement: [Eagle Tower, relativeSizeWithinCastle, one of the largest towers of Caernarfon Castle]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: relativeSizeWithinCastle Context triple: [Eagle Tower, relativeSizeWithinCastle, one of the largest towers of Caernarfon Castle]
-
A.
positionWithinCastle
Indicates that one entity is located inside the spatial boundaries of a castle.
-
B.
eraOfMainCastle
Indicates the historical period or era during which the main castle associated with an entity was built, used, or most prominent.
-
C.
relativeSizeWithinSite
chosen
Indicates the comparative size relationship of one entity to another within the same site or local context.
-
D.
relativeSizeWithinAirport
Indicates the comparative size relationship between entities specifically within the context or boundaries of an airport.
-
E.
nearbyCastle
Indicates that one entity is located close to or within a short distance of a castle.
- 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_69d6aa5d8be481909a43218b2bfdbe95 |
completed | April 8, 2026, 7:19 p.m. |
| NER | Named-entity recognition | batch_69d70fca498881909b40163a138cfb98 |
completed | April 9, 2026, 2:32 a.m. |
| PD | Predicate disambiguation | batch_69d6f309a44881908e49e3ba478c35b4 |
completed | April 9, 2026, 12:30 a.m. |
Created at: April 8, 2026, 9:14 p.m.