W09-3806 have trained the classifiers on a corpus matching a text entry application . However
S13-1042 instances from the full evaluation corpus matching words in the refined lexica .
N12-1044 entries on the English side yields a corpus matching Japanese transliterations with
P02-1012 pronominalization algorithms using corpus matching criteria instead of " quality
P02-1012 animate and all anaphora using a corpus matching technique . Furthermore , this
W11-2824 0.041 ¯ 6 An entry in the corpus matching one of the substitutions ( e.g.
hide detail