Triple
T18087997
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Mitral valve |
E432888
|
entity |
| Predicate | bestHeardAt |
P99033
|
FINISHED |
| Object | cardiac apex |
—
|
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: cardiac apex | Statement: [Mitral valve, bestHeardAt, cardiac apex]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: bestHeardAt Context triple: [Mitral valve, bestHeardAt, cardiac apex]
-
A.
heardBy
Indicates that a sound, message, or communication produced by one entity is perceived or received auditorily by another entity.
-
B.
isHeardIn
chosen
Indicates that a sound, voice, or audio from one entity is perceived or audible within the context, location, or medium of another entity.
-
C.
bestSeenIn
Indicates that something is most effectively or appropriately experienced, observed, or appreciated within a particular context, medium, or setting.
-
D.
typicallyHears
Indicates that one entity is the kind of thing that normally or characteristically hears another entity.
-
E.
mayHear
Indicates that one entity is permitted or able to hear sounds, speech, or audio produced 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_69d8b907d05c819083cc3bd6021089e6 |
completed | April 10, 2026, 8:47 a.m. |
| NER | Named-entity recognition | batch_69e4dd16234c8190b547e893a829d6c5 |
completed | April 19, 2026, 1:48 p.m. |
| PD | Predicate disambiguation | batch_69e4330e1f2881908b2506d47c48736b |
completed | April 19, 2026, 1:42 a.m. |
Created at: April 10, 2026, 10:27 a.m.