Triple
T4288265
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Fürth Hardhöhe |
E97323
|
entity |
| Predicate | locatedInDistrict |
P40
|
FINISHED |
| Object | Hardhöhe |
E97323
|
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: Hardhöhe | Statement: [Fürth Hardhöhe, locatedInDistrict, Hardhöhe]
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Hardhöhe Context triple: [Fürth Hardhöhe, locatedInDistrict, Hardhöhe]
-
A.
Hoche
Hoche is a Paris Métro station located in the northeastern suburb of Pantin, serving as a stop on the city’s Line 5.
-
B.
Hohe Marter
Hohe Marter is a subway station on the U3 line of the Nuremberg U-Bahn in Nuremberg, Germany.
-
C.
Hoch
Hoch is the nickname of Ed Hochuli, a well-known former NFL referee recognized for his muscular physique and detailed penalty explanations.
-
D.
Fürth Hardhöhe
chosen
Fürth Hardhöhe is a station in the city of Fürth that serves as the western terminus of a line on the Nuremberg U-Bahn rapid transit system.
-
E.
Rockheights
Rockheights is a residential neighbourhood in the municipality of Esquimalt on southern Vancouver Island in British Columbia, Canada.
- 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_69b3454595848190a0e6bbb6a2bea040 |
completed | March 12, 2026, 10:59 p.m. |
| NER | Named-entity recognition | batch_69b35060b70081908479c95b2afe8ec5 |
completed | March 12, 2026, 11:46 p.m. |
| NED1 | Entity disambiguation (via context triple) | batch_69b5c72bec388190b03048507e2e6858 |
completed | March 14, 2026, 8:38 p.m. |
Created at: March 12, 2026, 11:08 p.m.