Triple
T1362339
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | North Channel |
E29123
|
entity |
| Predicate | freezesDuring |
P19069
|
FINISHED |
| Object | winter |
—
|
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: winter | Statement: [North Channel, freezesDuring, winter]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: freezesDuring Context triple: [North Channel, freezesDuring, winter]
-
A.
freezesOver
chosen
Indicates that a liquid surface becomes solid due to low temperatures, typically forming a layer of ice over it.
-
B.
frozenIn
Indicates that one entity is immobilized or preserved in a solid, frozen state within or by another entity.
-
C.
frozenFrom
Indicates that one entity has been preserved or immobilized by being frozen starting from another entity, source, or prior state.
-
D.
requiresFrostFreeSeason
Indicates that the subject depends on a period without frost to grow, develop, or function properly.
-
E.
breaksDownWhen
Indicates that one entity becomes nonfunctional, damaged, or fails when subjected to the presence, action, or conditions imposed by another entity.
- 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_69a498d77abc8190913bf57e5f51d2c4 |
completed | March 1, 2026, 7:51 p.m. |
| NER | Named-entity recognition | batch_69a4c2b4ab3c8190ad692e32eee05976 |
completed | March 1, 2026, 10:50 p.m. |
| PD | Predicate disambiguation | batch_69a4bef945c08190a027472fdd695ea5 |
completed | March 1, 2026, 10:34 p.m. |
Created at: March 1, 2026, 7:57 p.m.