Hi, for Collins parser there is a wrapper script that generates the right format needed by Moses. This is described here: http://www.statmt.org/moses/?n=Moses.ExternalTools#collins
The format is described in the Syntax Tutorial: http://www.statmt.org/moses/?n=Moses.SyntaxTutorial There is currently no wrapper for any Chinese parser. The Stanford parser may do Chinese, but I have not tried it myself. http://nlp.stanford.edu/software/lex-parser.shtml -phi 2010/5/27 dongxinghua0213 <[email protected]>: > Hello, > I want to training syntax model using moses,but only Chinese-English > corpus are available(not labeled),So I download Collins parser ,but it > requires pos labeled corpus,then I download MXPOST to label the corpus > ,but the output is formatted as follows: > The_DT king_NN raised_VBD his_PRP$ general_JJ to_TO the_DT peerage_NN for_IN > winning_VBG the_DT > but the format collins parser requires seems like this , > The DT king NN raised VBD his_ RP$ general JJ to TO the DT peerage NN for IN > winning VBG the DT > to parse the English sentences in corpus, Could I leave out the "_" ? > Are there parser available for Chinese to train moses ? > > ________________________________ > 网易为中小企业免费提供企业邮箱（自主域名） > _______________________________________________ > Moses-support mailing list > [email protected] > http://mailman.mit.edu/mailman/listinfo/moses-support > >
_______________________________________________ Moses-support mailing list [email protected] http://mailman.mit.edu/mailman/listinfo/moses-support
