<DIV>Hi!</DIV>
<DIV> I am trying to develop a tagger for a highly agglutinating language known as Manipuri, falls under Tibeto Burman class of language. I want to develop a POS tagger for this language. There is no previously developed tagger for this language. I also think that for a language like Manipuri, morphological analysis will be an important task for assigning POS category to each word in a corpus. which one of the two approach - rule based or stochastic based will be appropriate for my task. </DIV>
<DIV> Please let me know your comments. If stochastic approach is appropriate please advice me how to proceed. </DIV>
<DIV> </DIV>
<DIV>with regards,</DIV>
<DIV>Siraj</DIV>
<DIV> </DIV><p><font face=arial size=-1>
<a href="http://in.mobile.yahoo.com" target="_blank"><b>Yahoo! India Mobile</a>:</b> Ringtones, Wallpapers, Picture Messages and more.
<font face=arial size=-1>Download <a href="http://in.mobile.yahoo.com" target="_blank"><b>now</a></b>.</font>