Triple
T20301022
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Pascal Soriot |
E505478
|
entity |
| Predicate | previousEmployer |
P1910
|
FINISHED |
| Object | Roche |
—
|
NE NERFINISHED |
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: Roche | Statement: [Pascal Soriot, previousEmployer, Roche]
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Roche Context triple: [Pascal Soriot, previousEmployer, Roche]
-
A.
Roche
Roche is a common surname of French origin borne by various notable individuals across fields such as architecture, politics, and the arts.
-
B.
Roche
chosen
Roche is a major Swiss multinational healthcare company and one of the world’s leading pharmaceutical and diagnostics firms.
-
C.
Novartis
Novartis is a global Swiss-based pharmaceutical company known for developing innovative medicines across a wide range of therapeutic areas.
-
D.
Sandoz
Sandoz is a historic Swiss pharmaceutical company best known as a predecessor of Novartis and a major player in generic medicines.
-
E.
Sanofi
Sanofi is a major French multinational pharmaceutical company known for developing prescription medicines, vaccines, and consumer healthcare products worldwide.
- F. None of above.
- G. Unsure - the case is ambiguous/there is not enough information to decide.
Provenance (2 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_69e0b4b8ab648190906e18538c250148 |
completed | April 16, 2026, 10:06 a.m. |
| NER | Named-entity recognition | batch_69e6770c700c81909da247cef0d0f1eb |
completed | April 20, 2026, 6:57 p.m. |
Created at: April 16, 2026, 11:17 a.m.