Triple
T3345994
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Pillars of Hercules |
E70375
|
entity |
| Predicate | flanks |
P20134
|
FINISHED |
| Object | western entrance of the Mediterranean Sea |
—
|
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: western entrance of the Mediterranean Sea | Statement: [Pillars of Hercules, flanks, western entrance of the Mediterranean Sea]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: flanks Context triple: [Pillars of Hercules, flanks, western entrance of the Mediterranean Sea]
-
A.
liesOnFlankOf
chosen
Indicates that one entity is positioned along the side or lateral edge of another entity.
-
B.
frontLine
Indicates that an entity is positioned at or associated with the foremost or primary line of engagement, activity, or defense relative to others.
-
C.
frontLineBetween
Indicates that a boundary or line of direct confrontation exists separating two opposing sides or regions.
-
D.
front
Indicates that one entity is located directly before or facing another entity along a primary viewing or movement direction.
-
E.
hasFlanker
Indicates that an entity is positioned or assigned as a flanking support or protection alongside 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_69ad85a405e48190b6e68de7cf9f319e |
completed | March 8, 2026, 2:20 p.m. |
| NER | Named-entity recognition | batch_69adb1f4ff888190bf14b9b7fbe9bcee |
completed | March 8, 2026, 5:29 p.m. |
| PD | Predicate disambiguation | batch_69ada42df1d48190874bb05f95deefde |
completed | March 8, 2026, 4:30 p.m. |
Created at: March 8, 2026, 3:12 p.m.