Triple
T420754
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Battle of Tarawa |
E8095
|
entity |
| Predicate | JapaneseCasualtiesKilled |
P6773
|
FINISHED |
| Object | over 4,000 |
—
|
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: over 4,000 | Statement: [Battle of Tarawa, JapaneseCasualtiesKilled, over 4,000]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: JapaneseCasualtiesKilled Context triple: [Battle of Tarawa, JapaneseCasualtiesKilled, over 4,000]
-
A.
casualtiesJapan
Indicates that an event or action resulted in casualties (deaths and/or injuries) occurring in Japan.
-
B.
militaryCasualtiesEstimate
chosen
Indicates an estimated number of people killed, wounded, or missing as a result of military conflict or operations.
-
C.
casualtiesKilledUS
Indicates that the relationship specifies the number of U.S. individuals who were killed as casualties in an event or incident.
-
D.
involvedShipJapan
Indicates that a ship associated with Japan was involved in the referenced event or activity.
-
E.
numberOfJewsKilled
Indicates the quantity of Jewish people who were killed in a given event, context, or time period.
- 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_69a2e7f1d1bc81909cf2dc9754a3c334 |
completed | Feb. 28, 2026, 1:04 p.m. |
| NER | Named-entity recognition | batch_69a2eebf65cc81908196fee4f09d5889 |
completed | Feb. 28, 2026, 1:33 p.m. |
| PD | Predicate disambiguation | batch_69a2edd3b948819097d96c73d0a0f699 |
completed | Feb. 28, 2026, 1:29 p.m. |
Created at: Feb. 28, 2026, 1:11 p.m.