language model 0.00390663
word language 0.0036288600000000002
word model 0.0035730700000000002
pos language 0.002836697
language models 0.0027191740000000004
model probability 0.0026358099999999997
target language 0.002597868
model approach 0.002560448
model probabilities 0.002509495
word models 0.0023856140000000003
surface language 0.002385349
language modeling 0.0023813280000000003
model figure 0.0023780299999999997
local language 0.002363482
many language 0.002348282
model reordering 0.002328651
factored language 0.0023148920000000003
complex language 0.0023106620000000002
local model 0.002307692
language pairs 0.002303699
global language 0.002284042
individual language 0.0022737110000000003
distortion model 0.002263191
language mod 0.0022585070000000003
supertag language 0.002238856
cached language 0.002238856
guage model 0.002209483
model applications 0.002199757
model adjectives 0.002183241
model corre 0.002183241
translation models 0.0021453740000000002
word position 0.002062982
phrase translation 0.0020298440000000003
word alignment 0.002015952
language 0.00198121
test data 0.001978628
translation system 0.00197535
model 0.00192542
word occurrence 0.0019116110000000001
computing word 0.0019065150000000001
training data 0.001869787
machine translation 0.001856141
english data 0.00183798
translation error 0.001805371
current translation 0.0018018679999999999
reference translation 0.0017376029999999999
translation systems 0.001715884
translation quality 0.001715102
translation sys 0.001679731
chine translation 0.001672287
translation candidates 0.001666842
translation hypoth 0.00166563
pos information 0.001648831
data sets 0.00164824
web data 0.001618332
linguistic data 0.0016023440000000001
pos models 0.001593451
annotated data 0.001533971
data conditions 0.001533971
data consor 0.001533971
pos tags 0.001457994
standard pos 0.0014240770000000002
translation 0.00140741
test set 0.001320305
pos tag 0.0013169
previous pos 0.00130936
speech recognition 0.001289529
lexical information 0.0012491619999999998
sentence bleu 0.001245796
target phrase 0.001239092
local pos 0.0012377590000000002
new approach 0.001227692
english sentence 0.001214111
different genres 0.00121167
pos lan 0.0011964130000000001
pos tagging 0.001181402
source phrase 0.0011747060000000002
words 0.0011646
global pos 0.001158319
side pos 0.0011471670000000002
likely pos 0.001143663
surface models 0.001142103
pos mod 0.001132784
pos tagger 0.001130493
joint probability 0.0011282319999999998
local models 0.0011202360000000001
target tokens 0.001104665
same feature 0.0011008720000000001
correlation results 0.0010860610000000001
test sets 0.001076268
distortion models 0.001075735
conditional probability 0.001073637
multiple test 0.00107256
ical information 0.001059139
positional information 0.001057084
baseline system 0.001053172
modeling approach 0.001035146
baseline decoder 0.001032659
individual models 0.001030465
guage models 0.001022027
