Triple
T26708172
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | butcher birds |
E673335
|
entity |
| Predicate | riskFromHumans |
P136373
|
FINISHED |
| Object | persecution as pest species |
—
|
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: persecution as pest species | Statement: [butcher birds, riskFromHumans, persecution as pest species]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: riskFromHumans Context triple: [butcher birds, riskFromHumans, persecution as pest species]
-
A.
threatToHumans
Indicates that the subject poses or represents a potential danger, harm, or risk to humans.
-
B.
riskElement
Indicates that one entity is a risk-related component, factor, or contributor associated with another entity within a risk context.
-
C.
riskToUser
Indicates that something poses a potential danger, harm, or adverse impact to the user.
-
D.
hasRiskFrom
chosen
Indicates that one entity is exposed to or may suffer potential harm, loss, or adverse effects as a result of another entity.
-
E.
riskTaken
Indicates that an entity has undertaken an action or decision involving exposure to potential loss, harm, or uncertainty.
- 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_69eecda3a22881908f3061c760b9d542 |
completed | April 27, 2026, 2:44 a.m. |
| NER | Named-entity recognition | batch_69f61a17a7788190946f7e32d63cd43f |
completed | May 2, 2026, 3:36 p.m. |
| PD | Predicate disambiguation | batch_69f611ab768c8190b1849c15a3e59dda |
completed | May 2, 2026, 3 p.m. |
Created at: April 27, 2026, 3:34 a.m.