Triple
T38378324
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Khost–Ghulam Khan road |
E893686
|
entity |
| Predicate | terminusInAfghanistan |
P78866
|
FINISHED |
| Object | Khost Province |
—
|
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: Khost Province | Statement: [Khost–Ghulam Khan road, terminusInAfghanistan, Khost Province]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: terminusInAfghanistan Context triple: [Khost–Ghulam Khan road, terminusInAfghanistan, Khost Province]
-
A.
terminusB_country
chosen
Indicates that the country specified is the location of the endpoint (terminus B) of a route, connection, or infrastructure segment.
-
B.
terminusA_country
Indicates the country in which the starting terminus (terminus A) of a route, line, or connection is located.
-
C.
closestMajorSettlementAfghanistan
Indicates the nearest major Afghan settlement (such as a large town or city) to a given location.
-
D.
terminusCountryNorth
Indicates the country in which the northern terminus (end point) of something, such as a route or line, is located.
-
E.
terminus
Indicates that one entity serves as the final endpoint or stopping place for another entity’s movement, route, or process.
- 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_69f76e4b1f748190a380696a16eae4a2 |
completed | May 3, 2026, 3:48 p.m. |
| NER | Named-entity recognition | batch_69fcec5f8b448190b48330a19b462d24 |
completed | May 7, 2026, 7:47 p.m. |
| PD | Predicate disambiguation | batch_69fceaf1e23881908ca24160a638e329 |
completed | May 7, 2026, 7:41 p.m. |
Created at: May 3, 2026, 4:31 p.m.