MST parser (maximum spanning tree dependency parser by R. McDonald) is used to determine the topology of a-layer trees. This is its reimplementation in Perl, with simplified MIRA algorithm (single-best MIRA is used).
Settings are provided via a config file accompanying the model file. The script loads the model
model_dir/model_name.model and its config <model_dir/model_name.config>.
You must set a model to use the parser, e.g.
model_name=en/conll_2007_best (if the default model dir
data/models/parser/mst_perl suits you; otherwise, also set
model_dir to a directory in which you have downloaded the models from
http://ufallab.ms.mff.cuni.cz/tectomt/share/data/models/mst_perl_parser/ or obtained in another way.)
It is not sensible to change the config file unless you decide to train your own model. However if you do decide to train your own model, then see Treex::Tool::Parser::MSTperl::Config.
TODO: provide a treex interface for the trainer?
Treex::Block::W2A::BaseChunkParser base clase (see the
Treex::Block::W2A::MarkChunks this block can be used before parsing to improve the performance by marking chunks (phrases) that are supposed to form a (dependency) subtree
Rudolf Rosa <email@example.com>
Copyright © 2011 by Institute of Formal and Applied Linguistics, Charles University in Prague
This module is free software; you can redistribute it and/or modify it under the same terms as Perl itself.
1 POD Error
The following errors were encountered while parsing the POD:
- Around line 259:
Non-ASCII character seen before =encoding in '©'. Assuming UTF-8