translation model 0.00451188
target phrase 0.0037427499999999995
word phrases 0.00363891
target word 0.0035322599999999997
translation rules 0.0035199100000000002
phrase pair 0.003510325
phrase length 0.003352385
phrase context 0.0032486719999999998
different word 0.0032295830000000003
translation models 0.003216657
phrase pairs 0.0032102019999999997
same phrase 0.0032080119999999997
translation system 0.003203798
phrase size 0.003188436
phrase extraction 0.003169219
phrase table 0.003166716
smt phrase 0.003152181
phrase clustering 0.003106969
clustering phrase 0.003106969
statistical phrase 0.0030865569999999998
labeling phrase 0.0030723549999999997
phrase labeling 0.0030723549999999997
phrase labels 0.00306748
noun phrase 0.003052633
hierarchical phrase 0.0030428029999999997
context word 0.003038182
initial phrase 0.003035085
phrase type 0.0030171679999999998
whole phrase 0.003005665
phrase boundary 0.0029995679999999998
translation performance 0.0029944
phrase types 0.002982335
phrase contexts 0.0029721319999999997
phrase instances 0.0029691049999999997
phrase sizes 0.00296298
first word 0.002961337
translation task 0.002939581
word alignments 0.002929632
machine translation 0.002896688
word clustering 0.002896479
unsupervised word 0.002885981
multiple word 0.002870224
word class 0.002863295
word clusters 0.002855203
translation systems 0.002855161
translation quality 0.002850832
only word 0.002836477
word tags 0.002833977
specific translation 0.002824385
reordering model 0.0028074230000000003
translation tasks 0.002804136
trained word 0.002789354
boundary word 0.002789078
word classes 0.002777644
vocabulary word 0.002772148
word types 0.002771845
distributional word 0.00276309
close translation 0.002760703
chine translation 0.002758987
word clusterings 0.002747906
ond word 0.002747625
ferred word 0.002746614
timate word 0.002746614
phrase 0.00271268
language model 0.002709443
bilingual model 0.002666901
translation 0.00251329
syntax model 0.002401723
clustering model 0.002392879
based model 0.002358597
model parameters 0.002330351
markov model 0.0023272650000000002
ing model 0.002316952
cal model 0.002292405
labeled model 0.0022852790000000003
bigram model 0.002268449
guage model 0.002254664
model variant 0.002242768
aware model 0.002242768
unaware model 0.002242768
training data 0.00221137
target phrases 0.00216679
model 0.00199859
source words 0.001935862
training sentence 0.001820505
target sentence 0.001791885
target language 0.001740923
different rules 0.001734013
source sentence 0.001724677
data set 0.001715435
other features 0.001681456
source language 0.001673715
parallel training 0.0016623279999999998
bilingual features 0.001642281
source length 0.001602567
feature vector 0.001577363
feature space 0.001571196
sentence pair 0.00155946
same feature 0.001556392
similar phrases 0.001549599
