Triple
T9195832
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Rachel Watson |
E220703
|
entity |
| Predicate | recoversMemoriesOf |
P51091
|
FINISHED |
| Object | Tom Watson’s abuse |
—
|
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: Tom Watson’s abuse | Statement: [Rachel Watson, recoversMemoriesOf, Tom Watson’s abuse]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: recoversMemoriesOf Context triple: [Rachel Watson, recoversMemoriesOf, Tom Watson’s abuse]
-
A.
rememberedThrough
Indicates that one entity is remembered, commemorated, or kept in memory by means of, or through the influence of, another entity or medium.
-
B.
recoveredIn
Indicates that something lost, damaged, or impaired has been restored or regained within a particular context, process, or location.
-
C.
recalledBy
chosen
Indicates that an entity is brought back, retrieved, or remembered as a result of another entity’s action or prompting.
-
D.
memorizedBy
Indicates that some content, information, or material has been learned and retained in memory by a particular entity.
-
E.
recovery
Indicates the process or state in which an entity regains a previous condition, function, or status after loss, damage, or decline.
- 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_69ca83e7ba70819088b74866d9da2c30 |
completed | March 30, 2026, 2:08 p.m. |
| NER | Named-entity recognition | batch_69ccd87c3a3c8190b60f19873ef6e1f8 |
completed | April 1, 2026, 8:34 a.m. |
| PD | Predicate disambiguation | batch_69cc660af2408190ae06eb8326e1c64e |
completed | April 1, 2026, 12:25 a.m. |
Created at: March 30, 2026, 7:25 p.m.