W dniu 2014-07-07 16:25, Marco A.G.Pinto pisze: > Hello! > > This is still the *"o"*->*"os"* and *"os"*->*"o"* issue which is > generating too many false positives. > > I was thinking about a way of adding possible exceptions to the XML in > order to fix most of the positives. > > I contacted my Minho University friends and they replied: > *"**the correct way of achieving that would be to have access to a > **morphologic **analyser and accessing the Hunspell dictionary, if it > was possible for the Hunspell dictionary to have such information > regarding the number of substantives/adjectives.**"* > > Daniel Naber told me this morning: > *"I suggest you have a look at the hunspell files to see if the > information is in there. If it's not, there might be other resources on > the internet with this information. "* > > Can someone help me?
Actually, you might try to see what results you get with our Galician module. I believe it comes from FreeLing, and Freeling also has a big Portuguese (probably Brazilian) analyzer we could reuse. But you might start your experiments with Galician. See: http://nlp.lsi.upc.edu/freeling/index.php?option=com_content&task=view&id=23&Itemid=58 Regards, Marcin ------------------------------------------------------------------------------ Open source business process management suite built on Java and Eclipse Turn processes into business applications with Bonita BPM Community Edition Quickly connect people, data, and systems into organized workflows Winner of BOSSIE, CODIE, OW2 and Gartner awards http://p.sf.net/sfu/Bonitasoft _______________________________________________ Languagetool-devel mailing list Languagetool-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/languagetool-devel
