Triple
T15189576
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Louis-Dreyfus family |
E362968
|
entity |
| Predicate | businessHoldings |
P117057
|
FINISHED |
| Object | agricultural commodities |
—
|
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: agricultural commodities | Statement: [Louis-Dreyfus family, businessHoldings, agricultural commodities]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: businessHoldings Context triple: [Louis-Dreyfus family, businessHoldings, agricultural commodities]
-
A.
businessBase
Indicates that one entity serves as the primary business foundation, core location, or main operational base for another entity.
-
B.
investsIn
Indicates that one entity allocates resources, typically money or capital, into another entity with the expectation of future returns or benefits.
-
C.
stocks
Indicates that one entity keeps or maintains a supply of another entity available for use or sale.
-
D.
spectrumHoldings
Indicates a relationship where an entity possesses or controls rights to use specific portions of the electromagnetic spectrum.
-
E.
foundedAsHoldingCompanyFor
Indicates that an entity was originally established specifically to serve as a holding company for another entity or group of entities.
- F. None of above. chosen
Provenance (4 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_69d85a09a39c81908759f23268e2d408 |
completed | April 10, 2026, 2:01 a.m. |
| NER | Named-entity recognition | batch_69e0067beedc8190abc0a94c7a38f85e |
completed | April 15, 2026, 9:43 p.m. |
| PD | Predicate disambiguation | batch_69deb97bd8bc8190b2ad4888f97cf963 |
completed | April 14, 2026, 10:02 p.m. |
| PDg | Predicate description generation | batch_69dec72059c08190a34f513a00185b08 |
completed | April 14, 2026, 11 p.m. |
Created at: April 10, 2026, 3:10 a.m.