You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
lexeme-internal structure to match UD tokenization (added subtokens in regular trees)
something for control/raising? cf. UD xcomp
finer-grained functions as used in CGEL/SIEG itself, for the benefit of readers (e.g. "Predicate", "Nucleus", "CatComp")
notes indicating the name of a special construction/phenomenon, CGEL page references, ambiguity, etc.
token and constituent offsets for computational use
In my view one of the main reasons to have a CGEL treebank (and eventually parser) is for explanation to humans, so the more pointers we have the better. But it's good to have a simple version of the tree that avoids clutter.
The text was updated successfully, but these errors were encountered:
This is a long-term idea: I'm starting to wonder if we should have an enhanced version of trees that might incorporate additional info such as
punctuation(added in regular trees)lemmas where different from wordform(added in regular trees)lexeme-internal structure to match UD tokenization(added subtokens in regular trees)xcomp
In my view one of the main reasons to have a CGEL treebank (and eventually parser) is for explanation to humans, so the more pointers we have the better. But it's good to have a simple version of the tree that avoids clutter.
The text was updated successfully, but these errors were encountered: