Triple
T36998607
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Icelandic Civil Protection |
E915292
|
entity |
| Predicate | typeOfHazardsManaged |
P1950
|
FINISHED |
| Object | natural disasters |
—
|
LITERAL 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: natural disasters | Statement: [Icelandic Civil Protection, typeOfHazardsManaged, natural disasters]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: typeOfHazardsManaged Context triple: [Icelandic Civil Protection, typeOfHazardsManaged, natural disasters]
-
A.
hasHazardManagement
Indicates that an entity has measures, systems, or responsibilities in place to control, mitigate, or respond to identified hazards.
-
B.
hazardType
chosen
Indicates the specific kind or category of hazard associated with an entity or situation.
-
C.
hazardManagementConcern
Indicates a relationship where one entity is a source, subject, or focus of concern regarding the identification, assessment, or control of hazards by another entity.
-
D.
hazardScope
Indicates the range or extent within which a particular hazard is relevant, applicable, or has effect.
-
E.
isHazardTo
Indicates that one entity poses a potential source of danger, harm, or risk to another entity.
- 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_69f76e8f1a8c81909db172ed31304971 |
completed | May 3, 2026, 3:49 p.m. |
| NER | Named-entity recognition | batch_6a03433dd64c8190ab620709153b6278 |
completed | May 12, 2026, 3:11 p.m. |
| PD | Predicate disambiguation | batch_6a034107e53881909127bac98cac3429 |
completed | May 12, 2026, 3:02 p.m. |
Created at: May 3, 2026, 4:14 p.m.