Linguistics Stack Exchange is a question and answer site for professional linguists and others with an interest in linguistic research and theory. It only takes a minute to sign up.
Sign up to join this community
Is there any straightforward way of identifying transitive verbs (or sentences containing transitive constructions) in an BrE English text?
I've looked into semantic shallow parsers, such as Semafor, but these don't seem to offer the straightforward approach I'm looking for.
Parsers for rich formalisms like C&C and Enju come with super-taggers. Super-taggers are quite like part-of-speech (POS) taggers, except that the tags they produce have more information.
For example, C&C produces:
word="undergone" lemma="undergo" pos="VBN" chunk="I-VP" cat="(S[pt]\NP)/NP"
The cat tells you that this is a verb that takes an NP on the right, and then one on the left to form an S.
However, things are not always this straight-forward. For "added constantly to the list", we get:
word="added" lemma="add" pos="VBN" chunk="I-VP" cat="(S[pt]\NP)/PP"
word="constantly" lemma="constantly" pos="RB" chunk="I-ADVP" cat="(S\NP)\(S\NP)"
word="to" lemma="to" pos="TO" chunk="I-PP" cat="PP/NP"
word="the" lemma="the" pos="DT" chunk="I-NP" cat="NP[nb]/N"
word="list" lemma="list" pos="NN" chunk="I-NP" cat="N"
"added", here is transitive, but one of its arguments is a PP. However, not all PPs fill valency slots.
I have no idea if you would find it more straight-forward than Semafor, or less. Boxer would give you discourse representation structures (DRSes) which include the kind of info of the Semafor output. I suppose you'll have to evaluate it for yourself.
