word alignment 0.00571777
word translation 0.00515092
alignment model 0.00436865
target word 0.004108558
word align 0.003914259
source word 0.003903437
first word 0.003896422
statistical word 0.003761872
head word 0.003727819
cal word 0.0037166789999999996
word sense 0.003694252
final word 0.0036566819999999996
word aligner 0.0036447299999999997
word classes 0.003644443
account word 0.003640943
perform word 0.003640943
word aligners 0.003640943
alignment models 0.0032551859999999998
alignment target 0.003127108
alignment results 0.003079195
alignment result 0.002850354
alignment links 0.002838692
alignment errors 0.0027807189999999996
alignment link 0.002780345
available alignment 0.0027636889999999997
cludes alignment 0.0026590879999999996
training data 0.002634179
translation probability 0.0025830000000000002
supervised model 0.002522573
unsupervised model 0.002474626
unlabeled data 0.002450886
ibm model 0.002449287
interpolation model 0.00243428
alignment 0.00236816
labeled data 0.002322364
interpolated model 0.002294617
polated model 0.002291974
machine translation 0.002216844
available data 0.002170069
model 0.00200049
translation 0.00180131
target sentence 0.001686318
target words 0.001677445
bilingual sentence 0.001663551
sentence pair 0.001497967
source sentence 0.001481197
learning algorithm 0.0014526959999999998
training set 0.001447799
estimation method 0.001375093
sentence pairs 0.0013701149999999999
supervised learning 0.00135489
ibm models 0.0013358229999999999
first method 0.001335307
source language 0.0013069330000000001
head words 0.001296706
training process 0.001257499
weighted training 0.0012498000000000001
ing method 0.0012209159999999998
ith sentence 0.0012190909999999998
entire training 0.001166683
fertility probability 0.001135756
distortion probability 0.001129542
second method 0.001125065
interpolation weight 0.001099931
parameter estimation 0.001086972
experimental results 0.001084478
nation method 0.001082917
tility probability 0.001073786
ment results 0.0010593080000000001
bilingual corpora 0.0010499110000000002
same way 0.001035995
final results 0.001018107
mental results 0.001003062
adaboost algorithm 9.70431E-4
reference set 9.35129E-4
sentence 9.2737E-4
words 9.18497E-4
error links 9.06982E-4
models 8.87026E-4
large amounts 8.756E-4
supervised adaboost 8.72625E-4
relative error 8.60822E-4
training 8.59639E-4
first round 8.414080000000001E-4
learning 8.32807E-4
position index 8.31437E-4
statistical machine 8.27796E-4
unsupervised adaboost 8.24678E-4
linear interpolation 8.19909E-4
ing methods 8.18554E-4
supervised boost 8.16885E-4
reference sets 8.1236E-4
single ensemble 8.026660000000001E-4
nlp tasks 8.00244E-4
error rate 7.948639999999999E-4
performance 7.94152E-4
method 7.88495E-4
entity classification 7.82317E-4
probability 7.8169E-4
possible solution 7.728570000000001E-4
