Constituency/Phrase Parsing
- is the task of breaking text/sentences into constituents (sub-phrases and words)
- non-terminals in the parse tree are types of phrases
- terminals are the words in the sentence
- builds on top of PoS tagging by combining the tags into larger constituents/phrases
- constituency parsing with Context-Free Grammar (CFG) or Phrase Structure Grammar (PSG) refers to the task of assigning proper trees to input strings/sentences
- proper here means a tree:
- that covers all and only elements of the input
- has an
Sat the top
- proper doesn’t actually mean that the tree is correct
- proper here means a tree:
- involves search which involves the making of choices
Constituency Parsing - Methods
|
Constituency Parsing Method |
Time Complexity (sentence of length 𝑛) |
|---|---|
|
𝑂(𝐶𝑛) where:
| |
|
𝑂(𝑛3·|𝐺|) where:
| |
|
𝑂(𝐾·𝑛3) where:
|