Triple

T4098517
Position Surface form Disambiguated ID Type / Status
Subject Helsinki Metro E87880 entity
Predicate depot P14646 FINISHED
Object Roihupelto depot
Roihupelto depot is a maintenance and storage facility serving the Helsinki Metro system in Helsinki, Finland.
 E412711 NE FINISHED

How this triple was built (4 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: Roihupelto depot | Statement: [Helsinki Metro, depot, Roihupelto depot]
NED1 Entity disambiguation (via context triple) gpt-5-mini-2025-08-07 
Target entity: Roihupelto depot
Context triple: [Helsinki Metro, depot, Roihupelto depot]
  • A. Holtet depot
    Holtet depot is a tram depot in Oslo, Norway, serving as one of the main facilities for housing and maintaining the city's tram fleet.
  • B. Grefsen depot
    Grefsen depot is a major tram depot in Oslo, Norway, serving as a key maintenance and storage facility for the city’s tram network.
  • C. Carnide depot
    Carnide depot is a maintenance and storage facility serving the Lisbon Metro system in Lisbon, Portugal.
  • D. Optila
    Optila was a Hun bodyguard who, along with fellow conspirators, carried out the assassination of Western Roman Emperor Valentinian III in 455 AD.
  • E. Munkkiniemi
    Munkkiniemi is a residential district in western Helsinki, Finland, known for its seaside location, parks, and notable architecture.
  • F. None of above. chosen
  • G. Unsure - the case is ambiguous/there is not enough information to decide.
NEDg Description generation gpt-5.1
Instruction
Generate a one-sentence description of the target entity. 
You are given a context triple in the form (subject, predicate, object), where the object is the target entity. 
# Instructions
Use the triple to infer relevant information about the entity. Describe the entity based on what is most defining, well-known. 
Avoid repeating the information from the triple, unless really essential.
# Response Format
Return only the sentence: "Description: [one-sentence description of the target entity]"
Input
Entity: Roihupelto depot
Triple: [Helsinki Metro, depot, Roihupelto depot]
Generated description
Roihupelto depot is a maintenance and storage facility serving the Helsinki Metro system in Helsinki, Finland.
NED2 Entity disambiguation (via description) gpt-5-mini-2025-08-07 
Target entity: Roihupelto depot
Target entity description: Roihupelto depot is a maintenance and storage facility serving the Helsinki Metro system in Helsinki, Finland.
  • A. Holtet depot
    Holtet depot is a tram depot in Oslo, Norway, serving as one of the main facilities for housing and maintaining the city's tram fleet.
  • B. Grefsen depot
    Grefsen depot is a major tram depot in Oslo, Norway, serving as a key maintenance and storage facility for the city’s tram network.
  • C. Carnide depot
    Carnide depot is a maintenance and storage facility serving the Lisbon Metro system in Lisbon, Portugal.
  • D. Optila
    Optila was a Hun bodyguard who, along with fellow conspirators, carried out the assassination of Western Roman Emperor Valentinian III in 455 AD.
  • E. Munkkiniemi
    Munkkiniemi is a residential district in western Helsinki, Finland, known for its seaside location, parks, and notable architecture.
  • F. None of above. chosen

Provenance (5 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_69aed94564cc8190a9c1457daedb6e7f completed March 9, 2026, 2:29 p.m.
NER Named-entity recognition batch_69aefd0bdea48190805a79515ee92709 completed March 9, 2026, 5:02 p.m.
NED1 Entity disambiguation (via context triple) batch_69b56b725788819091c6aaeccfb86964 completed March 14, 2026, 2:06 p.m.
NEDg Description generation batch_69b56c5bb12881908cf2c74d68ecd5b6 completed March 14, 2026, 2:10 p.m.
NED2 Entity disambiguation (via description) batch_69b56cc5c704819083dac59bf7b3cb83 completed March 14, 2026, 2:12 p.m.
Created at: March 9, 2026, 3:40 p.m.