word sequence 0.00255557
word length 0.002480883
order word 0.002414933
standard word 0.002371689
function word 0.002354505
word markov 0.002218
word chain 0.002201753
word sequences 0.002180221
explicit word 0.002163001
average word 0.00211881
particular word 0.002114509
word endings 0.002081594
word frequencies 0.002081594
proper word 0.002081594
training words 0.0019933859999999998
words kernel 0.001811882
feature space 0.001608997
feature vectors 0.001577961
size training 0.001560368
author model 0.001548804
sequence kernel 0.0015169720000000001
implicit feature 0.001417504
svm training 0.001385103
specific model 0.0013755360000000001
ing words 0.001374648
training dataset 0.001342225
training phase 0.001332618
such information 0.001329706
sequence length 0.001322353
sparse feature 0.001319359
kernel function 0.0013159069999999998
thor model 0.001313541
training material 0.001307437
feature elimination 0.001306114
recursive feature 0.0013021789999999999
sive feature 0.001295952
semantic features 0.001292504
full words 0.0012800350000000001
character sequence 0.001278062
individual words 0.001276415
related words 0.001256131
kernel approach 0.001250104
vidual words 0.001245213
cases features 0.0012317040000000001
combined training 0.00122613
discriminative features 0.001206276
approaches kernel 0.001199089
kernel approaches 0.001199089
essential features 0.001179176
kernel types 0.001147137
order character 0.001137425
sequence approach 0.001130172
size test 0.00110512
sequence kernels 0.001099153
model 0.00108164
quence kernel 0.001076924
feature 0.00107118
sequence analysis 0.00106808
order figure 0.0010612899999999999
different authors 0.0010578060000000001
symmetric kernel 0.001047682
length weight 0.001034394
other author 0.001020224
order probability 0.001011073
training 9.99956E-4
words 9.9343E-4
support vector 9.85836E-4
good performance 9.698319999999999E-4
probabilistic models 9.64683E-4
first experiment 9.644020000000001E-4
corresponding sequence 9.586950000000001E-4
space representation 9.53055E-4
sequence ker 9.517659999999999E-4
features 9.51284E-4
performance comparison 9.510059999999999E-4
ing size 9.4163E-4
character markov 9.40492E-4
empirical performance 9.40354E-4
ter sequence 9.35384E-4
text material 9.353689999999999E-4
different datasets 9.327370000000001E-4
acter sequence 9.32289E-4
sequence lengths 9.2813E-4
different lengths 9.26219E-4
character chain 9.24245E-4
biological sequence 9.22621E-4
ular sequence 9.22621E-4
chunk size 9.20634E-4
probability smoothing 9.18939E-4
order markov 9.18833E-4
specific length 9.177289999999999E-4
common threshold 9.1731E-4
other authors 9.142569999999999E-4
discrimination performance 9.104289999999999E-4
unmasking method 9.09668E-4
weight function 9.08016E-4
impostor text 9.062899999999999E-4
order chain 9.02586E-4
previous experiment 9.011640000000001E-4
final results 8.94432E-4
