Triple
T2197937
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Pranhita |
E50418
|
entity |
| Predicate | hasRiverIsland |
P970
|
FINISHED |
| Object | local sandbars and islands in dry season |
—
|
LITERAL 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: local sandbars and islands in dry season | Statement: [Pranhita, hasRiverIsland, local sandbars and islands in dry season]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: hasRiverIsland Context triple: [Pranhita, hasRiverIsland, local sandbars and islands in dry season]
-
A.
hasIsland
chosen
Indicates that one entity possesses, contains, or includes an island as part of its domain, territory, or structure.
-
B.
hasRiver
Indicates that a location or area contains, is traversed by, or is directly associated with a river.
-
C.
hasIslandAtMouth
Indicates that a geographic feature, typically a river, has an island located at or near its mouth where it meets a larger body of water.
-
D.
isIsland
Indicates that the subject entity is classified as an island, i.e., a landmass surrounded by water.
-
E.
hasIslandOnOneSide
Indicates that one side of an object, area, or boundary is adjacent to or bordered by an island.
- 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_69a88b044ab48190add007487680f009 |
completed | March 4, 2026, 7:41 p.m. |
| NER | Named-entity recognition | batch_69abbf79f3e08190b56e9d7c0ff27237 |
completed | March 7, 2026, 6:02 a.m. |
| PD | Predicate disambiguation | batch_69abbda706f4819094de73e1d1d1f539 |
completed | March 7, 2026, 5:54 a.m. |
Created at: March 4, 2026, 7:46 p.m.