[Corpora-List] Thanks for the reply and a summary.

Yuanyong Wang wyy at cse.unsw.EDU.AU
Mon May 30 11:47:21 UTC 2005


          Dear list memebers, thanks for the replies I've received. Here I
give a summary of the replies received:


          My original question:

            ... I also tried a couple of Chinese POS
taggers and parsers, some are ok, but apparently I am expecting something
better. Could anyone kindly suggest me with some top Chinses POS taggers
and syntactic parsers? ...



          Reply from Richard Xiao:

The best Chinese POS tagger I have tried is ICTCLAS (the Chinese Lexical
Analysis System developed by the Institute of Computing Technologies, the
Chinese Academia, Beijing), which is available (source codes and binary
compiled for Windows) at
http://mtgroup.ict.ac.cn/~zhp/ICTCLAS/index.html
<http://mtgroup.ict.ac.cn/~zhp/ICTCLAS/index.html>

I haven't tried their Chinese parser extensively, but the output of the
online demo seems alright.
http://mtgroup.ict.ac.cn/parserform.html
<http://mtgroup.ict.ac.cn/parserform.html>



          Reply from Huihsin Tseng:

Have you tried these two?
Bikel's parser can parse Chinese, English, Korean, and Arabic. See:
http://www.cis.upenn.edu/~dbikel/software.html
Roger Levy also has a Chinese parser based on Dan Klein's parser. See:
http://www-nlp.stanford.edu/software/lex-parser.shtml
They are written in java.



         Regards
         Robin.



More information about the Corpora mailing list