Triple
T11249347
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Perfect Dark |
E266285
|
entity |
| Predicate | developer |
P73
|
FINISHED |
| Object | Rare |
E266293
|
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: Rare | Statement: [Perfect Dark, developer, Rare]
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Rare Context triple: [Perfect Dark, developer, Rare]
-
A.
Rare
chosen
Rare is a renowned British video game developer known for creating influential titles such as Donkey Kong Country, GoldenEye 007, Banjo-Kazooie, and Sea of Thieves.
-
B.
Rare
Rare is Selena Gomez's third studio album, a pop record known for its themes of self-acceptance and emotional resilience.
-
C.
Hard to Find
"Hard to Find" is a reflective, melancholic indie rock song by The National, featured as the closing track on their album "Trouble Will Find Me."
-
D.
Few
Few is an English-language surname borne by various notable individuals, including American politician and Founding Father William Few.
-
E.
RAR
RAR is the Royal Australian Regiment, the principal regular infantry regiment of the Australian Army known for its service in major conflicts since World War II.
- 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_69d6aac7953c8190b82caf9d7640fdf9 |
completed | April 8, 2026, 7:21 p.m. |
| NER | Named-entity recognition | batch_69d7e93212448190b46b3799b0bdfa0f |
completed | April 9, 2026, 6 p.m. |
| NED1 | Entity disambiguation (via context triple) | batch_69e5257dd1e48190a9352fc8b62418da |
completed | April 19, 2026, 6:57 p.m. |
Created at: April 8, 2026, 9:31 p.m.