Triple
T24752222
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Zuiderzee |
E619174
|
entity |
| Predicate | partlyReclaimedAs |
P157291
|
FINISHED |
| Object | Eastern Flevoland |
—
|
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: Eastern Flevoland | Statement: [Zuiderzee, partlyReclaimedAs, Eastern Flevoland]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: partlyReclaimedAs Context triple: [Zuiderzee, partlyReclaimedAs, Eastern Flevoland]
-
A.
partlyReclaimedAs
chosen
Indicates that one entity has been partially converted, restored, or taken back into the state, use, or condition represented by the other entity.
-
B.
partiallyReusedAs
Indicates that one entity is used again as part of another entity, but only to a limited or incomplete extent rather than in its entirety.
-
C.
partlyRebuilt
Indicates that an entity has been reconstructed or restored to some extent, but not fully.
-
D.
partiallyReusable
Indicates that an entity can be reused to some extent but not fully or indefinitely, implying limited or partial reusability.
-
E.
partiallyRecapturedBy
Indicates that an entity that was previously captured or controlled has been regained only in part by another entity, not fully restored to its prior captured state.
- 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_69e2fabb349881908a13a212a0221a63 |
completed | April 18, 2026, 3:30 a.m. |
| NER | Named-entity recognition | batch_69f44a417a58819081777e18dda149fd |
completed | May 1, 2026, 6:37 a.m. |
| PD | Predicate disambiguation | batch_69f442a977b08190b44eac040cb90211 |
completed | May 1, 2026, 6:05 a.m. |
Created at: April 18, 2026, 4:25 a.m.