This project is in very initial stage.
“CharniakChineseParser” is a Charniak parser for Chinese. It is used for study purpose only.
See mytest.sh
only English now, output example:
I can’t love you.
(S1 (S (NP (PRP I)) (VP (MD can) (RB n’t) (VP (VB love) (NP (PRP you)))) (. .)))
(S1 (S (NP (PRP I)) (VP (MD ca) (RB n’t) (VP (VB love) (NP (PRP you)))) (. .)))
It is famous and fantastic English parser developed by Eugene Charniak. (with Mark Johnson?).
The reason use Charniak Parser as a base to develop Chinese :
- It’s speedy! 10 times faster than some slow parsers.
- It doesn’t use seperate “POS tagger” (no difference with Rule parser)
- It’s lexicalized. make it accurate.
- Code quality is rather high. (I have heavy comments for most code)
- Lack of document. Especially about detail of features.
- Features selection and model generation for Chinese。
There are several versions of Charniak Parser. This one is a relative old one (without ReRank) However its code size is also relatively small.