On Saturday, 2 April 2011, Mikel Forcada <m...@dlsi.ua.es> wrote: > > Dear Xiajian, > > I think you haven't taken your time to study how HMMs are used in > Apertium to perform part-of-speech disambiguation. Your proposal is > completely wrong as it confuses HMMs as used in speech recognition
Sorry to contradict, but the phoneme based approaches are valid. The problem is, these are based on the considerations required for Asian languages and not for European languages. > with HMMs as used for part-of-speech tagging. Make sure you write a > proposal that shows that you have studied (a) Apertium (by > installing it, studying its PoS tagger, etc.); (b) the documentation > available about Apertium part-of-speech tagging; (c) the problem of > "hidden" unknown words, and that makes a proposal to detect these > with an indication on how you plan to evaluate its performance (for > instance, by modifying existing dictionaries for a language pair to > induce unknowns). Modifying dictionaries would be the least efficient of the approaches I can imagine. -- <Leftmost> jimregan, that's because deep inside you, you are evil. <Leftmost> Also not-so-deep inside you. ------------------------------------------------------------------------------ Create and publish websites with WebMatrix Use the most popular FREE web apps or write code yourself; WebMatrix provides all the features you need to develop and publish your website. http://p.sf.net/sfu/ms-webmatrix-sf _______________________________________________ Apertium-stuff mailing list Apertium-stuff@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/apertium-stuff
