Triple

T3678852
Position Surface form Disambiguated ID Type / Status
Subject Patna E78060 entity
Predicate regionalLanguage P237 FINISHED
Object Magahi E53011 NE 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: Magahi | Statement: [Patna, regionalLanguage, Magahi]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07
Target entity: Magahi
Context triple: [Patna, regionalLanguage, Magahi]
  • A. Magahi chosen
    Magahi is an Eastern Indo-Aryan language spoken primarily in the Indian state of Bihar and surrounding regions.
  • B. Hastinapura
    Hastinapura is the ancient royal capital of the Kuru kingdom in Indian epic tradition, serving as the primary political and familial setting of the Mahabharata.
  • C. Tangail
    Tangail is a city in central Bangladesh known as a regional commercial hub and for its traditional handloom saree industry.
  • D. Kishkindha
    Kishkindha is the mythical monkey kingdom ruled by Sugriva in the Indian epic Ramayana, where Rama forms an alliance with the vanara army to search for Sita.
  • E. Mahoba
    Mahoba is a historic town in Uttar Pradesh, India, renowned for its role as a prominent center of power and culture in the Bundelkhand region, especially under the Chandela rulers.
  • F. None of above.
  • G. Unsure - the case is ambiguous/there is not enough information to decide.

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_69ad85e18c1c8190be8aafb227f39f48 completed March 8, 2026, 2:21 p.m.
NER Named-entity recognition batch_69adc46599188190a046eddb0d85c483 completed March 8, 2026, 6:48 p.m.
NED1 Entity disambiguation (via context triple) batch_69b4c3a9c6f48190a44b3b2dab37268d completed March 14, 2026, 2:10 a.m.
Created at: March 8, 2026, 3:25 p.m.