Triple
T5981442
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Donepezil |
E133125
|
entity |
| Predicate | seriousSideEffect |
P24552
|
FINISHED |
| Object | bradycardia |
—
|
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: bradycardia | Statement: [Donepezil, seriousSideEffect, bradycardia]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: seriousSideEffect Context triple: [Donepezil, seriousSideEffect, bradycardia]
-
A.
hasSeriousSideEffect
chosen
Indicates that an entity (such as a treatment, drug, or intervention) causes or is associated with a significant or severe adverse effect on another entity (typically a patient or biological system).
-
B.
possibleSideEffect
Indicates that one entity may occur as a side effect or unintended consequence of another entity or action.
-
C.
commonAdverseReactions
Indicates that the related entities are linked through adverse reactions or side effects that frequently occur in association with one another.
-
D.
hasCommonAdverseEffect
Indicates that two or more entities share at least one adverse effect that occurs in response to them.
-
E.
sideEffect
Indicates that one entity is an unintended or secondary effect resulting from the use or occurrence of 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_69c0086f45e8819098f73dd16d45ec9d |
completed | March 22, 2026, 3:19 p.m. |
| NER | Named-entity recognition | batch_69c04a6921b081908a6f6323d5c7a062 |
completed | March 22, 2026, 8 p.m. |
| PD | Predicate disambiguation | batch_69c049de98648190962b14fd341c93da |
completed | March 22, 2026, 7:58 p.m. |
Created at: March 22, 2026, 4:04 p.m.