Triple
T16174187
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Grenache-Syrah-Mourvèdre |
E392518
|
entity |
| Predicate | commonFoodPairing |
P14740
|
FINISHED |
| Object | grilled lamb |
—
|
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: grilled lamb | Statement: [Grenache-Syrah-Mourvèdre, commonFoodPairing, grilled lamb]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: commonFoodPairing Context triple: [Grenache-Syrah-Mourvèdre, commonFoodPairing, grilled lamb]
-
A.
typicalFoodPairing
chosen
Indicates that one food item is commonly served, consumed, or matched together with another as a customary or complementary pairing.
-
B.
foodInteraction
Indicates an interaction or relationship involving food between entities, such as consumption, sharing, preparation, or exchange.
-
C.
drinksWith
Indicates that two entities consume beverages together, typically at the same time and place in a social context.
-
D.
intendedFood
Indicates that one entity is the food item that another entity plans or is meant to eat or consume.
-
E.
usesIngredient
Indicates that one entity employs or incorporates another entity as an ingredient in its composition or creation.
- 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_69d87f1d32208190942e4e499a80c18c |
completed | April 10, 2026, 4:39 a.m. |
| NER | Named-entity recognition | batch_69e21eb9b8208190b60874cec7a3a98e |
completed | April 17, 2026, 11:51 a.m. |
| PD | Predicate disambiguation | batch_69e219d642708190ba31a90dce76a210 |
completed | April 17, 2026, 11:30 a.m. |
Created at: April 10, 2026, 5:02 a.m.