Triple
T22245470
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Plan G |
E549828
|
entity |
| Predicate | beneficiaryPays |
P12961
|
FINISHED |
| Object | Part B deductible out of pocket |
—
|
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: Part B deductible out of pocket | Statement: [Plan G, beneficiaryPays, Part B deductible out of pocket]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: beneficiaryPays Context triple: [Plan G, beneficiaryPays, Part B deductible out of pocket]
-
A.
beneficiaryContribution
Indicates that one party provides a contribution, support, or resources that benefit another designated beneficiary.
-
B.
paidBy
chosen
Indicates that a payment or financial obligation is made or settled by the specified entity.
-
C.
canHaveBeneficiary
Indicates that an entity is capable of having another party designated as its beneficiary, who may receive benefits, rights, or proceeds associated with it.
-
D.
beneficiaryCountry
Indicates that one country is the recipient or beneficiary of aid, resources, or advantages provided in a given context.
-
E.
beneficiarySelectionMethod
Indicates the way in which a beneficiary is chosen or determined in relation to an action, agreement, or resource.
- 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_69e11e41d9408190bd770cf282e22753 |
completed | April 16, 2026, 5:37 p.m. |
| NER | Named-entity recognition | batch_69f13217c9f88190aa2ce7d644b57739 |
completed | April 28, 2026, 10:18 p.m. |
| PD | Predicate disambiguation | batch_69e72fe1e0cc8190bd13cff2a0846225 |
completed | April 21, 2026, 8:05 a.m. |
Created at: April 16, 2026, 8:38 p.m.