Triple
T37061524
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | 琵琶湖 |
E917336
|
entity |
| Predicate | 主な流入河川 |
P57854
|
FINISHED |
| Object | 安曇川 |
—
|
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: 安曇川 | Statement: [琵琶湖, 主な流入河川, 安曇川]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: 主な流入河川 Context triple: [琵琶湖, 主な流入河川, 安曇川]
-
A.
largestInflowRiver
Indicates that one river has the greatest volume of water flowing into a specified body (such as a lake, sea, or another river) compared to all other contributing rivers.
-
B.
isMajorWatercourseOf
chosen
Indicates that a watercourse is a primary or significant river or stream associated with, or flowing through, a particular geographic area or feature.
-
C.
mainRiverAssociated
Indicates that one entity is associated with, or primarily linked to, a particular main river.
-
D.
flowsIntoViaMainRiver
Indicates that one water body ultimately flows into another through the course of a primary or main river channel.
-
E.
hasMajorRiverSystem
Indicates that an entity possesses or is traversed by a primary, large-scale river network that significantly characterizes its hydrology or geography.
- F. None of above.
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_69f76e95fa40819091e14681087ae5e4 |
completed | May 3, 2026, 3:49 p.m. |
| NER | Named-entity recognition | batch_69fb34e5576881909394355c8ec6ddd2 |
completed | May 6, 2026, 12:32 p.m. |
| PD | Predicate disambiguation | batch_69fb2f6171e88190bf1e0ee6a644b6a9 |
completed | May 6, 2026, 12:09 p.m. |
Created at: May 3, 2026, 4:14 p.m.