======================
The Statistical Parser
==============================================================================
This package provides cpp-implementation of the parser described in the paper.

To build the project, users can use the cmake command under "build" directory.

To run the parser, users should train the training-file's feature model and 
parse the test-file. 

More details can be found in commands. The following are the basic usage.

1: To train the 1ec_parsing models, use the following command.

  ./train.sh <train-file> <model-prefix> <train-iteration>

For example, users can train a toy parser using the sample file as follows.

  ./train.sh sample.conll08 sample 10

The model will be generated:
  sample.feat

2: To test the 1ec_parsing parser, use the following command.

  ./parse.sh <test-file> <model-file> <output-file>

For example, users can use the toy parser.

  ./parse.sh sample.conll08 sample.feat sample.out

The training/test files are in the CoNLL-2008 format. 

======================
The Statistical Parser
==============================================================================
We also include a sketch implementation (in both C++ and Python) of the two 
algoriths under the "coverage" directory. The functionality of the codes is 
to test whether it can recover a particular graph.

