Triple
T1922362
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | The Nutcracker (Balanchine version) |
E40150
|
entity |
| Predicate | costumeDesigner |
P184
|
FINISHED |
| Object | Karinska |
E44528
|
NE 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: Karinska | Statement: [The Nutcracker (Balanchine version), costumeDesigner, Karinska]
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Karinska Context triple: [The Nutcracker (Balanchine version), costumeDesigner, Karinska]
-
A.
Karinska
chosen
Karinska was a renowned 20th-century costume designer best known for her influential work in ballet and theater, particularly with the New York City Ballet.
-
B.
Seskarö
Seskarö is a Swedish island in the northern Baltic Sea known for its forests, beaches, and traditional fishing and forestry communities.
-
C.
Karlani
Karlani is one of the principal Pashtun tribal confederations, encompassing numerous tribes primarily located in eastern Afghanistan and northwestern Pakistan.
-
D.
Riasti
Riasti is a regional dialect of the Saraiki language spoken primarily in parts of southern Punjab, Pakistan.
-
E.
Kodava
Kodava is a Dravidian language spoken primarily by the Kodava (Coorg) community in the Kodagu district of Karnataka, India.
- F. None of above.
- G. Unsure - the case is ambiguous/there is not enough information to decide.
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_69a8864298748190a2f2fd34f7ef8d77 |
completed | March 4, 2026, 7:21 p.m. |
| NER | Named-entity recognition | batch_69abb233011881908736523d36f01b0c |
completed | March 7, 2026, 5:05 a.m. |
| NED1 | Entity disambiguation (via context triple) | batch_69adf3e458e8819098ea1c2d5598f890 |
completed | March 8, 2026, 10:10 p.m. |
Created at: March 4, 2026, 7:35 p.m.