Questionário vinculado ao guia · Médio Nível

Questionário de marcação de parte do discurso

Revise categorias de PDV, ambigüidade contextual, tags Penn Treebank, transições e emissões HMM, VBD, recursos de PNL, categorias UD e relações de sintaxe.

Caminhos de guia relacionadosMarcação de parte da fala
Pergunta 1 de 8

Which type of annotation does POS tagging assign to a token in context?

Visualize cada pergunta8 perguntas

As respostas e explicações aparecem conforme você responde ao teste.

  1. Which type of annotation does POS tagging assign to a token in context?
    • A document-level spam label only
    • A grammatical category such as noun, verb, or adjective
    • A translation into a required target language
    • A guarantee of the token's real-world referent
  2. Por que o contexto é essencial para a marcação de PDV?
    • Porque cada palavra tem exatamente uma tag possível
    • Porque a marcação ignora totalmente a frase
    • Porque frases mais longas são sempre mais fáceis
    • Porque muitas palavras podem ser diferentes classes gramaticais dependendo do uso
  3. How should the core Penn Treebank POS inventory be distinguished from its additional punctuation annotations?
    • It contains exactly two labels and no punctuation annotations
    • It is a database containing only book reviews
    • It has 36 core POS tags, with punctuation tags also used
    • It provides a different trained neural model for every tag
  4. Which probabilities does a classic HMM POS tagger combine when scoring a tag sequence?
    • Tag-to-tag transitions and word emissions given tags
    • Only word filenames and image brightness
    • Only probabilities of a tag given a word with no sequence model
    • A requirement that every tag have equal probability
  5. In the Penn Treebank POS convention, which grammatical property does VBD mark?
    • A singular noun
    • An adjective
    • A coordinating conjunction
    • A past-tense verb
  6. Which language-processing experiment can use POS tags as additional input features?
    • Compactação de arquivo de imagem
    • Reconhecimento de entidade nomeada
    • Audio amplifier volume control
    • Disk encryption
  7. How many universal POS categories does the stated Universal Dependencies version 2 inventory define?
    • Seventeen
    • Dois
    • Thirty-six core Penn categories
    • One hundred compulsory English-only categories
  8. In Universal Dependencies, how does NOUN differ from nsubj?
    • Both are required names for the same trained model
    • NOUN marks sentence sentiment; nsubj marks image size
    • NOUN is a word category; nsubj is a nominal-subject relation
    • Both necessarily mean that a word is a past-tense verb