Triple
T3039357
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Iraqi Jews |
E83086
|
entity |
| Predicate | maintain |
P39803
|
FINISHED |
| Object | distinct prayer melodies |
—
|
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: distinct prayer melodies | Statement: [Iraqi Jews, maintain, distinct prayer melodies]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: maintain Context triple: [Iraqi Jews, maintain, distinct prayer melodies]
-
A.
maintainsThrough
chosen
Indicates that one entity preserves, upholds, or continues a state, condition, or relationship by means of another specified entity, process, or medium.
-
B.
soughtToMaintain
Indicates an entity’s deliberate effort or intention to keep another entity or condition in its existing state over time.
-
C.
maintainedFor
Indicates that something is kept in a particular state, condition, or operation for the benefit, use, or purpose of another entity.
-
D.
maintenance
Indicates that an entity performs, requires, or is involved in upkeep, repair, or preservation activities for another entity or system.
-
E.
maintainedAs
Indicates that one entity is kept in a particular state, condition, or configuration by another entity through ongoing care, management, or support.
- 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_69ad8b2298908190a7cb4e9bdbf064d0 |
completed | March 8, 2026, 2:43 p.m. |
| NER | Named-entity recognition | batch_69ad9b2fa52c8190a7860f762d5232ab |
completed | March 8, 2026, 3:52 p.m. |
| PD | Predicate disambiguation | batch_69ad961fc62c819087c4c3a44b00847d |
completed | March 8, 2026, 3:30 p.m. |
Created at: March 8, 2026, 3:01 p.m.