Table 1.
Corpus description.
Table 2.
PAN/CLEF 2012 benchmark description.
Table 3.
Sentences in novels of BT.
Table 4.
Number of 3-grams per author (in thousands).
Table 5.
Novels of author BT.
Table 6.
Test and training sets of BT.
Table 7.
Distribution for training and test sets.
Fig 1.
Average accuracy (%) of SVM vs LR.
Table 8.
Accuracy (%) using ALL-features with different split settings.
Fig 2.
Average accuracy (%) of each type of n-gram for the authors.
Fig 3.
Average accuracy (%) for 3-grams using all features.
Table 9.
Accuracy (%) using PCA-features (1, 2, 3, 4 are split settings).
Fig 4.
Average accuracy (%) for 3-grams using PCA-features.
Table 10.
Accuracy using LSA-features (1, 2, 3, 4 are split settings).
Fig 5.
Average accuracy (%) of 3-grams using LSA features.
Fig 6.
Average of the accuracy (%) using all features.
Fig 7.
Average of the accuracy (%) using PCA features.
Fig 8.
Average of the accuracy (%) using LSA features.
Fig 9.
Accuracy (100%) obtained for different sizes.
Table 11.
Accuracy (%) using 3-grams in complete novels using ALL features.
Table 12.
Results for different types and sizes of n-grams in complete novels using ALL features.
Fig 10.
Accuracy (%) averages in models with and without a dimension reduction.
Table 13.
Precision, recall and F1 in ALL features.
Fig 11.
PCA visualization of Iris Murdoch (IM) using syntactic relationship 3-gram.
Fig 12.
PCA visualization of Louis Tracy (LT) using syntactic relationship 3-gram.
Fig 13.
ROC of Mark Twain (MT) using syntactic relationship 3-gram.
Fig 14.
AUC of Mark Twain (MT) using syntactic relationship 3-gram via different threshold.
Table 14.
One-sample T-test results for different types of 3-grams.