Triple
T37412999
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Phineas Nigellus Black |
E929622
|
entity |
| Predicate | laterAssists |
P121004
|
FINISHED |
| Object | Albus Dumbledore |
—
|
NE NERFINISHED |
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: Albus Dumbledore | Statement: [Phineas Nigellus Black, laterAssists, Albus Dumbledore]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: laterAssists Context triple: [Phineas Nigellus Black, laterAssists, Albus Dumbledore]
-
A.
assistsWith
Indicates that one entity helps or supports another entity in performing or accomplishing a specific task, activity, or objective.
-
B.
assistsOver
Indicates that one entity provides help or support to another entity in a way that surpasses or exceeds a certain reference level, standard, or counterpart.
-
C.
wasAssistedBy
Indicates that an entity carried out an action or achieved a result with the help or support of another entity.
-
D.
laterHelps
chosen
Indicates that one entity provides help or assistance to another at a subsequent time rather than immediately.
-
E.
winnerAssistsPerGameApprox
Indicates the approximate number of assists per game recorded by the winning participant or team.
- 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_69f76ebde49481908566cd96b37ccc84 |
completed | May 3, 2026, 3:50 p.m. |
| NER | Named-entity recognition | batch_6a011c64ff18819099195701bbecbb20 |
completed | May 11, 2026, 12:01 a.m. |
| PD | Predicate disambiguation | batch_6a011b7f7a508190b2ebb518fb7a2fe9 |
completed | May 10, 2026, 11:57 p.m. |
Created at: May 3, 2026, 4:16 p.m.