H05-1031 most important features for the classi cation task . For the major/minor classi
H05-1031 20-fold cross validation for both classi cation tasks . The results are shown
H05-1031 what features we used for our two classi cation tasks ( cf. list of features
C04-1035 categories described be - low . The classi cation was done by 3 expert annotators
C04-1035 part-of-speech tagging , or to dialogue act classi cation . We formulate our typing constraints
C04-1035 solid foundation for the current classi cation . It also indicates that it is
C04-1035 techniques to extract rules for sluice classi cation in dialogue . In Section 2 we
C04-1035 learning approach to bare sluice classi cation in dialogue using corpus-based
H01-1001 author The activities used for classi cation are those of the semi-naive coder
C04-1035 well suited to the task of sluice classi cation in dialogue on the basis of these
C04-1061 been mapped from the more complex classi cation in DiMLex under kat ( German
H01-1001 rithm . The features used for classi cation are words the 50 most frequent
C04-1113 also semantic change and verb classi cation , and to determine where irregularities
C04-1128 be the best contributor to the classi cation task . But , when the number
H05-1031 feature 13 ) . As in the minor-major classi cation , the syntactic choices for reference
C04-1133 corpus analysis for verb sense classi cation . 1.2 Word Sense Disambiguation
C04-1013 rarely available . Even in a binary classi cation problem , there will often be
C04-1128 described above . Ripper outputs a classi cation model for predicting the class
C04-1128 contribution of a single feature to the classi cation task may not be intuitively apparent
C04-1035 suited to the task of dialogue act classi cation for fragments on the basis of
hide detail