Triple
T2190621
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Remicade |
E49851
|
entity |
| Predicate | isContraindicatedIn |
P23156
|
FINISHED |
| Object | severe heart failure |
—
|
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: severe heart failure | Statement: [Remicade, isContraindicatedIn, severe heart failure]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: isContraindicatedIn Context triple: [Remicade, isContraindicatedIn, severe heart failure]
-
A.
hasContraindication
chosen
Indicates that one entity (such as a treatment, drug, or procedure) should not be used or performed in the presence of another entity (such as a condition, factor, or co-medication) because it may cause harm or adverse effects.
-
B.
notRecommendedBy
Indicates that one entity explicitly advises against or does not endorse another entity.
-
C.
restrictedFromUseFor
Indicates that something is prohibited or limited from being used for a specified purpose, context, or application.
-
D.
consistentWith
Indicates that one entity does not contradict and is compatible or in agreement with another entity, condition, or set of constraints.
-
E.
hasCommonAdverseEffect
Indicates that two or more entities share at least one adverse effect that occurs in response to them.
- 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_69a88aaba3c48190b351cab9b26989ff |
completed | March 4, 2026, 7:40 p.m. |
| NER | Named-entity recognition | batch_69abbf9e99f08190892d34485c8f2f25 |
completed | March 7, 2026, 6:03 a.m. |
| PD | Predicate disambiguation | batch_69abbda32d1881909d1fd83a751fb21c |
completed | March 7, 2026, 5:54 a.m. |
Created at: March 4, 2026, 7:46 p.m.