Triple
T3698368
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Andaman Islands |
E78511
|
entity |
| Predicate | contains |
P35
|
FINISHED |
| Object | Interview Island |
E149766
|
NE 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: Interview Island | Statement: [Andaman Islands, contains, Interview Island]
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Interview Island Context triple: [Andaman Islands, contains, Interview Island]
-
A.
Interview Island
chosen
Interview Island is a small, sparsely inhabited island in the Andaman archipelago of India, known for its wildlife and protected forest areas.
-
B.
Switch Island
Switch Island is a major road junction in Merseyside, England, where several key routes including motorways and A-roads converge.
-
C.
House Island
House Island is a small, privately owned island in Casco Bay, Maine, known for its historic Fort Scammel and scenic coastal views.
-
D.
The Island
The Island is a 2005 science fiction thriller film directed by Michael Bay that explores themes of human cloning, identity, and corporate ethics.
-
E.
The Island
The Island is a 1979 thriller novel by Peter Benchley that follows a journalist who uncovers a hidden community of modern-day pirates in the Caribbean.
- F. None of above.
- G. Unsure - the case is ambiguous/there is not enough information to decide.
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_69ad85e3b1888190abc983e06968696d |
completed | March 8, 2026, 2:21 p.m. |
| NER | Named-entity recognition | batch_69adc512ba188190a15bcacafac3f476 |
completed | March 8, 2026, 6:50 p.m. |
| NED1 | Entity disambiguation (via context triple) | batch_69b4c3dbb7fc81909df2e7a4755cce28 |
completed | March 14, 2026, 2:11 a.m. |
Created at: March 8, 2026, 3:26 p.m.