Triple
T19715497
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Bandarawela |
E473465
|
entity |
| Predicate | hasNearbyAttraction |
P2064
|
FINISHED |
| Object | Diyaluma Falls |
—
|
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: Diyaluma Falls | Statement: [Bandarawela, hasNearbyAttraction, Diyaluma Falls]
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Diyaluma Falls Context triple: [Bandarawela, hasNearbyAttraction, Diyaluma Falls]
-
A.
Diyaluma Falls
chosen
Diyaluma Falls is one of Sri Lanka’s tallest and most scenic waterfalls, renowned for its dramatic cascades and natural rock pools that attract many visitors.
-
B.
Boyoma Falls
Boyoma Falls is a series of powerful cataracts on the Lualaba River in the Democratic Republic of the Congo, known as one of the largest waterfall systems in Africa by volume.
-
C.
Tanda Falls
Tanda Falls is a scenic waterfall and popular natural getaway located near Mirzapur in Uttar Pradesh, India.
-
D.
Mooula Falls
Mooula Falls is a scenic waterfall located in the lush, remote Halawa Valley on the island of Molokai in Hawaii.
-
E.
Malanda Falls
Malanda Falls is a picturesque waterfall and popular swimming spot in Far North Queensland, Australia, known for its lush rainforest surroundings and easy accessibility.
- F. None of above.
- G. Unsure - the case is ambiguous/there is not enough information to decide.
Provenance (2 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_69d8e516dd048190a0b6c93ea3e71f58 |
completed | April 10, 2026, 11:55 a.m. |
| NER | Named-entity recognition | batch_69e6440cb47c81908124dfbd6f781d23 |
completed | April 20, 2026, 3:19 p.m. |
Created at: April 10, 2026, 1:46 p.m.