Triple
T4251159
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Flag of the British Virgin Islands |
E95854
|
entity |
| Predicate | lampArrangement |
P2362
|
FINISHED |
| Object | around Saint Ursula |
—
|
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: around Saint Ursula | Statement: [Flag of the British Virgin Islands, lampArrangement, around Saint Ursula]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: lampArrangement Context triple: [Flag of the British Virgin Islands, lampArrangement, around Saint Ursula]
-
A.
arrangementFeatures
Indicates that one entity specifies or characterizes the structural or compositional features of an arrangement involving another entity.
-
B.
arrangement
chosen
Indicates a relationship where entities are organized, ordered, or positioned in a particular configuration or sequence relative to one another.
-
C.
needleArrangement
Indicates how needles are positioned or organized relative to each other in a given context.
-
D.
litOn
Indicates that one entity is illuminated or activated by a light source associated with another entity.
-
E.
arrangementType
Indicates the specific kind or category of arrangement that characterizes how the related entities are organized or structured in relation to each other.
- 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_69b3453f759881909b91f01a1e82c036 |
completed | March 12, 2026, 10:59 p.m. |
| NER | Named-entity recognition | batch_69b34ea059108190aa00f0f20673e0ac |
completed | March 12, 2026, 11:39 p.m. |
| PD | Predicate disambiguation | batch_69b347f73e008190a908a48ef389945a |
completed | March 12, 2026, 11:10 p.m. |
Created at: March 12, 2026, 11:06 p.m.