Triple
T22634289
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Kailash Nath Katju |
E558637
|
entity |
| Predicate | officeStartPositionHeld |
P73495
|
FINISHED |
| Object | Governor of West Bengal, 1948 |
—
|
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: Governor of West Bengal, 1948 | Statement: [Kailash Nath Katju, officeStartPositionHeld, Governor of West Bengal, 1948]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: officeStartPositionHeld Context triple: [Kailash Nath Katju, officeStartPositionHeld, Governor of West Bengal, 1948]
-
A.
officeStartQualifier
Indicates a modifier that qualifies or specifies the circumstances, timing, or conditions under which an office term begins.
-
B.
officeStartFor
chosen
Indicates the time or date at which an entity begins its term, role, or period of service in an office or official position.
-
C.
officeStartAfter
Indicates that one entity’s term or period in office begins after another specified time or event.
-
D.
officeStartApproximate
Indicates that the start of a person’s term in office is known only approximately, not as an exact date.
-
E.
officeContestedHeldBy
Indicates that a particular office or position that was contested in an election is currently held by a specified 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_69e245467d9881908d6985bd0db7a1f1 |
completed | April 17, 2026, 2:35 p.m. |
| NER | Named-entity recognition | batch_69f1700be10c8190830393fdbec1033d |
completed | April 29, 2026, 2:42 a.m. |
| PD | Predicate disambiguation | batch_69ee62855558819080da946c7b35a160 |
completed | April 26, 2026, 7:07 p.m. |
Created at: April 17, 2026, 3:03 p.m.