Triple
T37114205
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Mumbai–Ahmedabad route |
E919073
|
entity |
| Predicate | connectsWaterBodyRegion |
P147860
|
FINISHED |
| Object | Arabian Sea coast |
—
|
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: Arabian Sea coast | Statement: [Mumbai–Ahmedabad route, connectsWaterBodyRegion, Arabian Sea coast]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: connectsWaterBodyRegion Context triple: [Mumbai–Ahmedabad route, connectsWaterBodyRegion, Arabian Sea coast]
-
A.
connectsBodyOfWater
Indicates a relationship where one entity serves as a link or passage between two bodies of water, allowing them to be joined or interact.
-
B.
connectsToBodyOfWaterRegion
chosen
Indicates that a region has a direct physical connection or boundary with a body of water.
-
C.
waterbodyRegion
Indicates that a water body is located within, associated with, or spans a particular geographic region.
-
D.
connectedBodyOfWater
Indicates that two geographic locations are linked by a continuous body of water through which water can flow between them.
-
E.
addressesWaterBody
Indicates that an action, plan, or measure is directed toward managing, affecting, or dealing with a specific water body.
- 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_69f76e9c57148190ba789dd059645bb9 |
completed | May 3, 2026, 3:49 p.m. |
| NER | Named-entity recognition | batch_69ffcb5536d88190bfc2e00b854cacfb |
completed | May 10, 2026, 12:03 a.m. |
| PD | Predicate disambiguation | batch_69ffc900c2a081909dea04aa60566923 |
completed | May 9, 2026, 11:53 p.m. |
Created at: May 3, 2026, 4:15 p.m.