Triple
T28315186
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | the second Mrs. de Winter |
E717113
|
entity |
| Predicate | feelsOvershadowedBy |
P29489
|
FINISHED |
| Object | Rebecca de Winter |
—
|
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: Rebecca de Winter | Statement: [the second Mrs. de Winter, feelsOvershadowedBy, Rebecca de Winter]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: feelsOvershadowedBy Context triple: [the second Mrs. de Winter, feelsOvershadowedBy, Rebecca de Winter]
-
A.
overshadowedBy
chosen
Indicates that one entity’s importance, visibility, or impact is diminished because another entity is more prominent or dominant.
-
B.
isOverlookedBy
Indicates that one entity has a view over or looks down upon another entity from a higher or vantage position.
-
C.
dominatedBy
Indicates that one entity is under the control, influence, or overpowering authority of another entity.
-
D.
towersOver
Indicates that one entity is significantly taller or higher in position than another, often in a way that appears dominant or imposing.
-
E.
canBeSubduedBy
Indicates that one entity is capable of overpowering, restraining, or otherwise bringing another entity under control.
- 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_69eff6e6c3b08190ad78de6ba7f04548 |
completed | April 27, 2026, 11:53 p.m. |
| NER | Named-entity recognition | batch_69f644e5e3c8819092a5295a8c566930 |
completed | May 2, 2026, 6:39 p.m. |
| PD | Predicate disambiguation | batch_69f641e0fde08190bf06a1c5b388aa84 |
completed | May 2, 2026, 6:26 p.m. |
Created at: April 28, 2026, 12:19 a.m.