Triple
T11978310
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | TAV Airports Holding |
E285091
|
entity |
| Predicate | hasSubsidiary |
P254
|
FINISHED |
| Object | Havaş |
E386835
|
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: Havaş | Statement: [TAV Airports Holding, hasSubsidiary, Havaş]
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Havaş Context triple: [TAV Airports Holding, hasSubsidiary, Havaş]
-
A.
Havaş
chosen
Havaş is a Turkish ground handling and airport services company operating at numerous airports domestically and internationally.
-
B.
Havza
Havza is a district and town in northern Turkey known for its thermal springs and location within Samsun Province in the Black Sea region.
-
C.
Kocaali
Kocaali is a coastal town and district in northwestern Turkey, situated along the Black Sea in Sakarya Province.
-
D.
Gürbulak
Gürbulak is a Turkish border village and crossing point on the frontier with Iran, serving as a key gateway between the two countries.
-
E.
Yassıada
Yassıada is one of Istanbul’s Princes' Islands in the Sea of Marmara, historically known for its use as a place of exile and for hosting the 1960–61 trials of Turkish political leaders.
- 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_69d6ab2eaeb881909f7914758f859413 |
completed | April 8, 2026, 7:23 p.m. |
| NER | Named-entity recognition | batch_69d90393cfb08190b5b45d3e5e32fad3 |
completed | April 10, 2026, 2:05 p.m. |
| NED1 | Entity disambiguation (via context triple) | batch_69f471f6afc48190856a0f7c486b28aa |
completed | May 1, 2026, 9:27 a.m. |
Created at: April 8, 2026, 9:46 p.m.