Triple
T1971174
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | ʻOheʻo Gulch |
E42801
|
entity |
| Predicate | nearestTown |
P350
|
FINISHED |
| Object | Hāna |
E214001
|
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: Hāna | Statement: [ʻOheʻo Gulch, nearestTown, Hāna]
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Hāna Context triple: [ʻOheʻo Gulch, nearestTown, Hāna]
-
A.
Hāna
chosen
Hāna is a small, remote town on the eastern coast of Maui, Hawaii, known for its lush landscapes, scenic coastal views, and the famously winding Road to Hāna.
-
B.
Kihei
Kihei is a coastal town on the southwest shore of Maui, Hawaii, known for its sunny weather, beaches, and resort and residential communities.
-
C.
Līhuʻe
Līhuʻe is the principal town and administrative center of Kauai County on the Hawaiian island of Kauai.
-
D.
Wailuku
Wailuku is a historic town on the Hawaiian island of Maui that serves as the county seat and a cultural and administrative center.
-
E.
Mililani
Mililani is a master-planned residential community and census-designated place located in the central part of the island of Oahu in Hawaii.
- 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_69a8871289048190b00b0d7744b7b2b1 |
completed | March 4, 2026, 7:25 p.m. |
| NER | Named-entity recognition | batch_69abb3d2836c8190a35cb6d8e2dd4bdf |
completed | March 7, 2026, 5:12 a.m. |
| NED1 | Entity disambiguation (via context triple) | batch_69ae95eb4db481909ab7507147f1d674 |
completed | March 9, 2026, 9:42 a.m. |
Created at: March 4, 2026, 7:36 p.m.