Triple
T23754244
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Beachy Head Lighthouse |
E587059
|
entity |
| Predicate | previousLightReplaced |
P64595
|
FINISHED |
| Object | Belle Tout Lighthouse |
—
|
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: Belle Tout Lighthouse | Statement: [Beachy Head Lighthouse, previousLightReplaced, Belle Tout Lighthouse]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: previousLightReplaced Context triple: [Beachy Head Lighthouse, previousLightReplaced, Belle Tout Lighthouse]
-
A.
replacedEarlierLight
Indicates that one light source has been substituted or superseded by another light source that existed or was installed earlier in time.
-
B.
previousLightStations
chosen
Indicates that one light station existed earlier in time or sequence than another related light station.
-
C.
laterLightSource
Indicates that one light source occurs or is activated after another in time.
-
D.
previousLocationReplaced
Indicates that one location has been superseded or taken over by another location, replacing its previous role or position.
-
E.
plannedReplacement
Indicates that one entity is intended or scheduled to be replaced by another entity in the future.
- 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_69e2490a0eec81908cdef8a862828d7a |
completed | April 17, 2026, 2:51 p.m. |
| NER | Named-entity recognition | batch_69f1bcc561b081908bfd516ef5e2dcfd |
completed | April 29, 2026, 8:09 a.m. |
| PD | Predicate disambiguation | batch_69f155f012808190a4b1cbc155558ade |
completed | April 29, 2026, 12:50 a.m. |
Created at: April 17, 2026, 7:13 p.m.