On Tue, Dec 28, 2010 at 12:42 AM, Shashwat Anand <anand.shash...@gmail.com> wrote: > Regarding dictionary lookup+n-gram approach I didn't quite understand what > you wanted to say.
Run through trigram analysis first, if it identified multiple languages as being matches within the error margin then split the text into words, and look up each word in the respective dictionaries to get a second opinion. Katie -- CoderStack http://www.coderstack.co.uk/python-jobs The Software Developer Job Board -- http://mail.python.org/mailman/listinfo/python-list
