Triple
T19064578
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | რაბათის ციხე |
E466619
|
entity |
| Predicate | nearbyCity |
P350
|
FINISHED |
| Object | ასპინძა |
—
|
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: ასპინძა | Statement: [რაბათის ციხე, nearbyCity, ასპინძა]
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: ასპინძა Context triple: [რაბათის ციხე, nearbyCity, ასპინძა]
-
A.
Aspindza
chosen
Aspindza is a small town in southern Georgia known for its proximity to the historic cave city of Vardzia and its location in the Samtskhe-Javakheti region.
-
B.
Aspa
Aspa is a small municipality located in the Segrià comarca within the province of Lleida in Catalonia, Spain.
-
C.
Asbarez
Asbarez is a long-running Armenian-American newspaper that serves as a key news source and voice for the Armenian diaspora, particularly in the United States.
-
D.
Aspet
Aspet is a small commune in southwestern France, situated in the foothills of the Pyrenees in the Haute-Garonne department.
-
E.
Aspronisi
Aspronisi is a small uninhabited volcanic islet in the Santorini (Thera) caldera in Greece, known for its steep cliffs and role in the ancient volcanic complex.
- 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_69d8dd040fb881909af2a964f65ad208 |
completed | April 10, 2026, 11:20 a.m. |
| NER | Named-entity recognition | batch_69e5e19799d8819099d0b848771ae425 |
completed | April 20, 2026, 8:19 a.m. |
Created at: April 10, 2026, 12:03 p.m.