Triple
T9882926
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Laysan albatross |
E180854
|
entity |
| Predicate | breedingSubstrate |
P8037
|
FINISHED |
| Object | sandy or grassy ground |
—
|
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: sandy or grassy ground | Statement: [Laysan albatross, breedingSubstrate, sandy or grassy ground]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: breedingSubstrate Context triple: [Laysan albatross, breedingSubstrate, sandy or grassy ground]
-
A.
aquariumTrade
Indicates that an entity is involved in the commercial trade or collection of organisms for keeping in aquariums.
-
B.
reproductiveHabitat
chosen
Indicates the type of environment or location where an organism carries out its reproductive activities or where its offspring develop.
-
C.
breedingUse
Indicates that one entity is used for breeding or reproductive purposes with respect to another entity.
-
D.
isSubstrateOf
Indicates that one entity serves as the material or reactant upon which another entity (typically an enzyme or process) acts in a chemical or biochemical reaction.
-
E.
breedingNote
Indicates notes or remarks documenting details, conditions, or observations related to a breeding event or process between entities.
- 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_69ca828082cc8190a40f8d299caa6545 |
completed | March 30, 2026, 2:02 p.m. |
| NER | Named-entity recognition | batch_69cdb418417c8190b00dea8916a2e102 |
completed | April 2, 2026, 12:11 a.m. |
| PD | Predicate disambiguation | batch_69cd1d810ed48190a252b70e9390c8f3 |
completed | April 1, 2026, 1:28 p.m. |
Created at: March 30, 2026, 8:38 p.m.