/conll2016

CoNLL 2016 Shallow Discourse Parsing System (Soochow University)

Primary LanguagePython

CoNLL2016 Shallow Discourse Parsing

Soochow University Shallow Discourse Parsing System

Components

  1. Connective Identification ==> connective.py

  2. Argument Labeling ==> argument.py

  3. Explicit Sense Classification ==> explicit.py

  4. NonExplicit Sense Classification ==> nonexp.py

Running

Training each model respectively

python connective.py

python argument.py

python explicit.py

python nonexp.py

Get end-to-end result

python end2end.py -o ../report -r dev.out.json

===== N.B. zh/ part is not the final submission system for chinese shallow discourse parsing.