MST parser adjusted to parsing English sentences. Just a lightweight wrapper for Treex::Block::W2A::ParseMSTperl which is base clase. This class just sets the English model
conll_2007_small as the default and defines a method for obtaining coarse-grained PoS tags (first two characters of PennTB-like tags).
The default model is very small and good only for testing and demonstration purposes. Use e.g.
model_name=conll_2007_best instead to get good results.
Copyright 2011 Rudolf Rosa This file is distributed under the GNU General Public License v2. See $TMT_ROOT/README.