Raphael Cohen - 2012-06-25

This is an implementation of EasyFirst parser with a precompiled model for the biomedical domain trained on the Genia Treebank.
The package is based on Yoav Goldberg's implementation of EasyFirst Parser.

It is based on POS tagging preprocessing using GeniaTagger by Tsuruoka et al.
Download and install separately

Usage

Generate the POS input

Usage: geniatagger [textFile] > [posTextFile]

Example: geniatagger file.txt > fileWithPOS.txt

-This should be run in the local directory where genia tagger is installed.

Run Parser

% cat fileWithPOS.txt | python transformGeniaPOStoConllx.py | python sdparser.py -conll -noextra

  • This will output the parsed sentences in conll format to the stdout (screen).
 

Last edit: Raphael Cohen 2012-06-25