Kedis ku xidhan hagaha · Dhexdhexaad Heerka

Soo Saaridda Xidhiidhka Kediska Qoraalka

Tijaabi sida soo saarista xidhiidhku uu qoraalka ugu beddelo xaqiiqooyin habaysan, iyo sida dhuumaha caadiga ah iyo hababka kormeerku u fashilmi karaan.

Wadooyinka haga ee la xidhiidhaKa Soo Saarista Xiriirka Qoraalka
Su'aal 1 ee 8

Waa maxay wax soo saarka caadiga ah ee nidaamka soo saarista xidhiidhka?

Horudhac su'aal kasta8 su'aalo

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

  1. Waa maxay wax soo saarka caadiga ah ee nidaamka soo saarista xidhiidhka?
    • Saddex jibbaar habaysan oo isku xidha labada hay'adood xidhiidh
    • Geed naxwe ahaan jumlada
    • Dhibcaha dareenka min taban ilaa togan
    • Turjumaadda jumlada oo loo beddelayo luqad kale
  2. In a common pipeline for relation extraction, what may identify entity mentions before pair classification?
    • Qoraal-ku-hadal-ku-habboon
    • Aqoonsiga cidda loo magacaabay
    • Qaybta sawirka
    • Audio source separation
  3. Why can distant supervision produce noisy relation-extraction labels?
    • Knowledge bases contain no structured facts
    • It requires every sentence to be hand-labeled
    • A sentence may mention two linked entities without stating their knowledge-base relation
    • It removes entity mentions before training
  4. Sidee badi moodooyinka RE-ga ku salaysan ee transformer-ku tilmaamayaa lamaanaha si loo kala saaro?
    • Adiga oo u turjumaya jumlada marka hore
    • Adigoo tirtiraya mid ka mid ah hay'adaha
    • Adigoo qoraalka u beddelaya maqal
    • Adiga oo gelinaya calaamado gaar ah oo ku wareegsan labada hay'adood
  5. Maxaa kala sooca xidhiidhka 'furan' ee soo saarista heerka RE?
    • Waxay si toos ah uga soo saartaa odhaahyada xidhiidhka qoraalka iyada oo aan la sii qeexin noocyada xidhiidhka
    • Waxay ku shaqaysaa oo kaliya tirooyinka
    • Waxay u baahan tahay xog calaamadaysan oo ka badan hab kasta oo kale
    • Waxay heli kartaa hal xiriir oo keliya dukumeenti kasta
  6. Why include a no-relation label when classifying candidate entity pairs?
    • To translate the sentence into another language
    • To represent pairs for which the text does not assert a target relation
    • To replace named entity recognition
    • To guarantee that every pair has a relation
  7. What does document-level relation extraction allow a system to use?
    • Only the spelling of a single entity
    • Audio features from a spoken recording
    • Evidence that may be spread across multiple sentences
    • A fixed relation without reading text
  8. How does open relation extraction differ from a closed relation schema?
    • It cannot use text as input
    • It always requires a knowledge base to label every sentence
    • It only identifies entity boundaries
    • It can express relation phrases without choosing only from a predefined label inventory