Triple
T24302266
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Gakh District |
E612429
|
entity |
| Predicate | hasNaturalBorder |
P155478
|
FINISHED |
| Object | Greater Caucasus Mountains |
—
|
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: Greater Caucasus Mountains | Statement: [Gakh District, hasNaturalBorder, Greater Caucasus Mountains]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: hasNaturalBorder Context triple: [Gakh District, hasNaturalBorder, Greater Caucasus Mountains]
-
A.
hasBorderThrough
Indicates that a border between two regions or entities passes through or along a specified intermediate area, feature, or object.
-
B.
haveOpenBorderUnder
Indicates that one entity maintains an open-border policy or regime while being under the jurisdiction, authority, or framework of another entity.
-
C.
hasWaterBorderWith
Indicates that two entities share a common boundary that runs along or across a body of water, such as a river, lake, or sea.
-
D.
hasSouthBorder
Indicates that one entity shares its southern boundary or border with another entity.
-
E.
hasBorderRelation
Indicates that one entity shares a boundary or border with another entity.
- F. None of above. chosen
Provenance (4 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_69e2d7d91bb48190bc5377d17a85fb21 |
completed | April 18, 2026, 1:01 a.m. |
| NER | Named-entity recognition | batch_69f29222d3008190af765d62787ee334 |
completed | April 29, 2026, 11:20 p.m. |
| PD | Predicate disambiguation | batch_69f1c45c6ec081908401b69424428100 |
completed | April 29, 2026, 8:42 a.m. |
| PDg | Predicate description generation | batch_69f1c6d4e99081909f61899eccafb73e |
completed | April 29, 2026, 8:52 a.m. |
Created at: April 18, 2026, 1:17 a.m.