Dependency Parsing Quiz

Answers and explanations appear as you take the quiz.

1. In dependency parsing, how is a sentence's structure represented?
   • As a tree of directed, labeled links from each word to its head
   • As nested phrase brackets only
   • As a flat list of word frequencies
   • As a spectrogram of the audio
2. In 'The dog chased the cat,' what is typically the root of the dependency tree?
   • dog
   • chased
   • the
   • cat
3. How does dependency parsing differ from constituency parsing?
   • It groups words into nested phrases rather than linking individual words
   • It only works on numbers
   • It links individual words by direct relations rather than grouping them into nested phrases
   • It ignores verbs entirely
4. What is an advantage of transition-based parsers?
   • They require no training data
   • They only work for English
   • They output audio
   • They are typically fast, running in linear time
5. What does Universal Dependencies offer researchers and parser developers?
   • Shared annotation guidelines and dependency relations used across many language treebanks
   • A dataset containing only English audio
   • A fixed list of named entities for every document
   • A guarantee that every parser reaches the same accuracy
6. What does a biaffine scorer do in a neural dependency parser?
   • It converts sentences directly into audio
   • It scores candidate head-dependent arcs and their dependency labels
   • It assigns image masks to pixels
   • It guarantees a fixed accuracy on every benchmark
7. In a basic dependency tree, what is true of a non-root word?
   • It must be the sentence’s main verb
   • It has no relation label
   • It has one syntactic head
   • It belongs to exactly one nested phrase only
8. How does a graph-based parser commonly form a dependency tree?
   • By translating each token independently
   • By grouping words only into phrase brackets
   • By removing all function words before parsing
   • By scoring candidate arcs and selecting a well-formed tree