word hmm 0.0026727
word error 0.002655697
word context 0.002546872
word level 0.0024314009999999997
tag features 0.002380097
pos features 0.002354565
word segment 0.002342922
lexical features 0.002340889
rent word 0.002318385
new features 0.002284895
features table 0.00227388
additional features 0.002195901
joint features 0.002132393
input features 0.0021159209999999998
conjunction features 0.002080331
ping features 0.002041579
hmm model 0.0020302700000000003
different feature 0.002012159
statistical model 0.0018595410000000001
parsing model 0.001848957
arabic sentence 0.001835737
features 0.00181921
entropy model 0.001802725
maxent model 0.0017988260000000001
fst model 0.001744873
arabic text 0.001743324
many feature 0.00173361
diacritization model 0.001731991
feature space 0.001714878
competitive model 0.0016897770000000002
mentation model 0.00168113
ent model 0.001680676
powerful model 0.001678091
feature interactions 0.00167432
feature sets 0.001673184
multiple feature 0.001662892
feature types 0.001659502
arabic segmentation 0.00165862
specific feature 0.001635063
interesting feature 0.001626991
feature categories 0.001617983
arabic tree 0.001572164
arabic speakers 0.001546073
arabic diacritics 0.001523334
arabic treebank 0.00152232
arabic letters 0.001494127
arabic scripts 0.001487862
modern arabic 0.001486672
arabic documents 0.001470223
arabic dialects 0.001453789
arabic master 0.001450433
arabic blank 0.001450433
arabic alphabet 0.001450433
model 0.00144281
several words 0.001435002
training data 0.001416146
feature 0.0013691
many words 0.00136653
same data 0.001335581
tag sequence 0.001294074
different system 0.001282334
arabic 0.00122725
calize words 0.001225234
other applications 0.001188972
training set 0.001187233
other approaches 0.001187177
syntactic information 0.001161324
training corpus 0.001149868
character sequence 0.001140925
such information 0.001135236
system performance 0.001122433
other diacritics 0.001121854
data split 0.001098538
testing data 0.001096764
treebank data 0.001095729
other techniques 0.001093549
such state 0.001092158
sequence classification 0.001091855
context information 0.001090177
other states 0.001074041
segmentation system 0.001070645
data the 0.001055782
diacritic sequence 0.001047996
morphological analyzer 0.001037605
analysis system 0.0010368740000000001
input sequence 0.001029898
speech recognition 0.001018144
markov sequence 0.0010161880000000001
language modeling 0.001012057
tagging system 0.001010503
words 0.00100202
state machine 9.90949E-4
entire sequence 9.890580000000001E-4
parse information 9.87316E-4
sequence informa 9.80412E-4
morphological units 9.78561E-4
markov dependency 9.7641E-4
statistical pos 9.52086E-4
based system 9.472440000000001E-4
different diacritics 9.391429999999999E-4
