This is an example script to run the code of my ACL 2014 paper Character-Level Chinese Dependency Parsing
Please follow the below steps:
(1) Download the code of ZPar.
(2) Modify the "setting" file, finding the following line
#chinese character-based dependency parser: jcad, jcadeager
ensuring that CHINESE_POSDEPENDENCY_PARSER jcad
(3) run "python configure.py chinese posdepparser"
(4) run command cmake . generate Makefile
(5) compile using make
(6) copy the bin files in the bin folder into current folder
(7) run "./ctb70dep-exp.sh" in current github (not ZPar folder)
================
One can use my word structure annotations to generate the training corpus.
By the way, if you have purchased the data of CTB, you could email me directly for the completed training data, my email is mason.zms@gmail.com.