Triple
T12820758
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | HAF |
E306522
|
entity |
| Predicate | abbreviation |
P43
|
FINISHED |
| Object | HAF |
E306522
|
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: HAF | Statement: [HAF, abbreviation, HAF]
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: HAF Context triple: [HAF, abbreviation, HAF]
-
A.
HAF
chosen
HAF is the commonly used abbreviation for the Hellenic Air Force, the air warfare branch of Greece’s armed forces.
-
B.
HFA
HFA is the IATA airport code for Haifa Airport, a regional airport serving the city of Haifa in northern Israel.
-
C.
HFA
HFA is the commonly used acronym for the New York State Housing Finance Agency, a public authority that finances affordable housing developments across New York State.
-
D.
HAJ
HAJ is the three-letter IATA airport code for Hannover Airport in Hanover, Germany.
-
E.
HAP
HAP is Apple's HomeKit Accessory Protocol, a communication standard that defines how smart home accessories securely interact with Apple devices and the Home app.
- 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_69d7bdf46c448190b1faa55aaacb6317 |
completed | April 9, 2026, 2:55 p.m. |
| NER | Named-entity recognition | batch_69d96e9e5a2481908921d097df541db8 |
completed | April 10, 2026, 9:41 p.m. |
| NED1 | Entity disambiguation (via context triple) | batch_69f68ed165188190a4cac781c753fb23 |
completed | May 2, 2026, 11:54 p.m. |
Created at: April 9, 2026, 5:31 p.m.