Since we deal with a rather diverse data set, we constructed
separate subjectivity classifiers for each of our product categories. We trained the classifier using a Dynamic Language
Model classifier with n-grams (n = 8) from the LingPipe toolkit4. The accuracy of the classifiers according to the Area
under the ROC curve, measured using 10-fold cross validation
was: 0.85 for audio and video players, 0.87 for digital cameras,
and 0.82 for DVDs.