Triple
T3837201
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Tourbillon Castle |
E91161
|
entity |
| Predicate | locatedIn |
P40
|
FINISHED |
| Object | Sion |
E14340
|
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: Sion | Statement: [Tourbillon Castle, locatedIn, Sion]
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Sion Context triple: [Tourbillon Castle, locatedIn, Sion]
-
A.
Sion
chosen
Sion is a historic Swiss town in the canton of Valais, known for its hilltop castles, vineyards, and role as a regional cultural and administrative center.
-
B.
Sirion
Sirion is one of the greatest rivers of Middle-earth in J.R.R. Tolkien’s legendarium, flowing from the northern regions of Beleriand to the sea and playing a central role in many events of the First Age.
-
C.
Valleiry
Valleiry is a small French commune in the Haute-Savoie department of the Auvergne-Rhône-Alpes region in southeastern France, near the Swiss border.
-
D.
Levasy
Levasy is a small city located in Jackson County in the U.S. state of Missouri.
-
E.
Yeola
Yeola is a town in the Nashik district of Maharashtra, India, known historically as the birthplace of the Indian freedom fighter Tatya Tope.
- 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_69aed960b538819096561c8ed448dec9 |
completed | March 9, 2026, 2:29 p.m. |
| NER | Named-entity recognition | batch_69aeeb9baa508190800e73bf186f046e |
completed | March 9, 2026, 3:47 p.m. |
| NED1 | Entity disambiguation (via context triple) | batch_69b50405264c8190b145dc3929ffc940 |
completed | March 14, 2026, 6:45 a.m. |
Created at: March 9, 2026, 3:18 p.m.