Thanks Jorn; I'm working on trying out the tokens right now, but I'll look at training.
--- On Mon, 2/13/12, Jörn Kottmann <kottm...@gmail.com> wrote: > From: Jörn Kottmann <kottm...@gmail.com> > Subject: Re: New to opennlp > To: opennlp-users@incubator.apache.org > Date: Monday, February 13, 2012, 6:12 PM > On 02/14/2012 12:07 AM, Michael > Collins wrote: > > Thanks for your response. > > > > Does opennlp provide a way to create the *.train file > based on a body of text which I provide, or is the *.train > file created another way. > > > > Can i use the SentenceDetectME.sentDetect() method > without providing a *.train file? > > > > > > > > Yes your first lines should do that: > java.io.FileInputStream ModelIn = new > java.io.FileInputStream(ModelPath > + "en-sent.zip"); > opennlp.tools.sentdetect.SentenceModel SentnModel = new > opennlp.tools.sentdetect.SentenceModel(ModelIn); > opennlp.tools.sentdetect.SentenceDetectorME SentenceDetectME > = new > opennlp.tools.sentdetect.SentenceDetectorME(SentnModel); > > Now you should be able to use the sentence detector with a > model > form our website. > > Depending on your use case and domain you might need to do > the training > yourself. > > The train file just contains one sentence per line. This way > the line > break is used to indicate the end of a sentence. > > Jörn >
