If I parse following sentence:
He is playing cricket in ground with grass.
with stanford parser, the result is:
(ROOT (S (NP (PRP He)) (VP (VBZ is) (VP (VBG playing) (NP (NP (NN cricket)) (PP (IN in) (NP (NN ground)))) (PP (IN with) (NP (NN grass))))) (. .)))
Is there any parser who can correct the result on the basis of probability as the probability of appearing grass with ground is higher than cricket?
Or parser which do chunking before parsing like this:
He is playing cricket |in ground| |with grass|.
and calculate probability with all combinations before generating the parse tree.
- He is playing cricket |in ground| |with grass|.
- He is playing cricket |with grass| |in ground|.