Triple
T58088
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | NBA Finals games |
E1149
|
entity |
| Predicate | tieBreakerForHomeCourt |
P635
|
FINISHED |
| Object | head-to-head record |
—
|
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: head-to-head record | Statement: [NBA Finals games, tieBreakerForHomeCourt, head-to-head record]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: tieBreakerForHomeCourt Context triple: [NBA Finals games, tieBreakerForHomeCourt, head-to-head record]
-
A.
hasAdvantage
chosen
Indicates that one entity possesses a benefit, edge, or favorable position over another in a given context.
-
B.
majorChallenger
Indicates that one entity is a primary competitor or leading rival to another in a given context or domain.
-
C.
homeFieldOpened
Indicates that a team's designated home field has officially begun operation or been opened for use.
-
D.
hasAutomaticPlayoffBids
Indicates that certain participants are guaranteed entry into a playoff competition without needing to qualify through regular selection or performance criteria.
-
E.
teamPlayedFor
Indicates that a person was a member of and played for a particular sports team.
- 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_69a248adc5b48190aa8db9fb092fb28a |
completed | Feb. 28, 2026, 1:45 a.m. |
| NER | Named-entity recognition | batch_69a24c9057348190aa6692eeeae19569 |
completed | Feb. 28, 2026, 2:01 a.m. |
| PD | Predicate disambiguation | batch_69a24ac7547c81909bb68f327cdb9158 |
completed | Feb. 28, 2026, 1:54 a.m. |
Created at: Feb. 28, 2026, 1:50 a.m.