Triple
T2328705
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Kuril–Kamchatka subduction system |
E48350
|
entity |
| Predicate | riskFactorFor |
P26874
|
FINISHED |
| Object | tsunami hazard in northwest Pacific |
—
|
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: tsunami hazard in northwest Pacific | Statement: [Kuril–Kamchatka subduction system, riskFactorFor, tsunami hazard in northwest Pacific]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: riskFactorFor Context triple: [Kuril–Kamchatka subduction system, riskFactorFor, tsunami hazard in northwest Pacific]
-
A.
riskFeature
Indicates that one entity possesses or exhibits a characteristic, condition, or attribute that increases the likelihood or severity of a negative outcome for another entity or situation.
-
B.
riskType
Indicates the category or nature of risk associated with an entity, event, or relationship.
-
C.
riskBasis
chosen
Indicates the underlying factor, condition, or rationale that forms the basis for assessing or assigning risk in a given context.
-
D.
riskGroup
Indicates that an entity belongs to a category of individuals or items that share an elevated level of risk relative to others.
-
E.
riskAddressed
Indicates that a particular risk has been identified and is being mitigated, managed, or otherwise handled by an associated action, control, or measure.
- 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_69a88aa308a88190b0b86c011fda7fce |
completed | March 4, 2026, 7:40 p.m. |
| NER | Named-entity recognition | batch_69abcc30c5e881908c5d526d7e7491d0 |
completed | March 7, 2026, 6:56 a.m. |
| PD | Predicate disambiguation | batch_69abc5926d048190a535e3f23d41de2a |
completed | March 7, 2026, 6:28 a.m. |
Created at: March 4, 2026, 7:50 p.m.