Triple
T33957432
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Georgian transport network |
E870617
|
entity |
| Predicate | linksToCountry |
P95796
|
FINISHED |
| Object | Turkey |
—
|
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: Turkey | Statement: [Georgian transport network, linksToCountry, Turkey]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: linksToCountry Context triple: [Georgian transport network, linksToCountry, Turkey]
-
A.
linksCountry
chosen
Indicates a relationship where one entity connects or associates another entity specifically with a country.
-
B.
relatedCountry
Indicates that there is a relevant or associated relationship between an entity and a specified country, without specifying the exact nature of that relationship.
-
C.
servesCountryConnections
Indicates that an entity provides or facilitates service connections between different countries.
-
D.
repositoryCountry
Indicates the country in which a repository is located or maintained.
-
E.
relatedCountryGroup
Indicates that there is an association or linkage between an entity and a specific group of countries.
- 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_69f3499c2d7481909c953a5010227725 |
completed | April 30, 2026, 12:22 p.m. |
| NER | Named-entity recognition | batch_69f7064e906881909c3186c646145d34 |
completed | May 3, 2026, 8:24 a.m. |
| PD | Predicate disambiguation | batch_69f70100ec1c8190a6b97f50e88891f2 |
completed | May 3, 2026, 8:02 a.m. |
Created at: May 1, 2026, 1:49 a.m.