Triple
T16142598
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | assassination of James A. Garfield |
E391696
|
entity |
| Predicate | victimTermStart |
P121271
|
FINISHED |
| Object | 1881-03-04 |
—
|
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: 1881-03-04 | Statement: [assassination of James A. Garfield, victimTermStart, 1881-03-04]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: victimTermStart Context triple: [assassination of James A. Garfield, victimTermStart, 1881-03-04]
-
A.
victimOfficeStartDate
Indicates the date on which the victim began their role, position, or office associated with the described event or relationship.
-
B.
victimStatus
Indicates the condition or state of a person who has been harmed or wronged as a result of an event, action, or offense.
-
C.
victimState
Indicates the condition or status that a victim is in as a result of an event, action, or harmful incident.
-
D.
victimRole
Indicates that one entity participates in an event or situation specifically in the role of the victim or harmed party.
-
E.
victimCountPolicy
Indicates the rule or criterion used to determine how victims are counted or classified in a given context.
- F. None of above. chosen
Provenance (4 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_69d87f1c65e48190aa2b4c472e9bafc4 |
completed | April 10, 2026, 4:39 a.m. |
| NER | Named-entity recognition | batch_69e21d9082588190bc7e6ff491f0d94e |
completed | April 17, 2026, 11:46 a.m. |
| PD | Predicate disambiguation | batch_69e182885bc08190822ae7e8a4b8ac1f |
completed | April 17, 2026, 12:44 a.m. |
| PDg | Predicate description generation | batch_69e1835b64948190ae1d2a9d4cc64acf |
completed | April 17, 2026, 12:48 a.m. |
Created at: April 10, 2026, 5:01 a.m.