translation model 0.00594526
phrase translation 0.0056036
translation models 0.004701639
log translation 0.004642669
translation search 0.004618968
translation score 0.00456486
translation probability 0.004530983
translation scores 0.00452872
training data 0.00451852
other translation 0.004506332
machine translation 0.004505113
translation problem 0.004500885
speech translation 0.00447731
reference translation 0.004463402
lexical translation 0.0044629809999999995
translation rule 0.00445992
translation probabilities 0.004445346
translation hypothesis 0.004444026
human translation 0.004429286
quality translation 0.004421247
translation quality 0.004421247
new translation 0.004383667
full translation 0.004362748
translation tokens 0.004354332
translation rules 0.004331149
unique translation 0.00432306
translation hypotheses 0.0043067949999999995
translation hypothe 0.004293582
extract translation 0.004287502
translation prob 0.004286161
gle translation 0.004283094
translation adap 0.004281671
translation 0.00399287
training method 0.0037780599999999998
training sentence 0.00370015
training corpus 0.00364926
training sentences 0.003432587
training set 0.003212359
parallel training 0.003182999
quality training 0.003003597
training experiments 0.002991026
web training 0.002990058
training procedure 0.00292993
training sen 0.002923841
training bitext 0.002920639
whole training 0.002904577
rate training 0.00290083
overall training 0.002873126
suitable training 0.002868924
lbfgs training 0.002866703
robust training 0.002866264
native training 0.002863659
tive training 0.0028627740000000002
language model 0.002858347
data weighting 0.002600324
training 0.00257522
parallel data 0.002551079
model adaptation 0.002544685
model score 0.00252438
model probability 0.002490503
space model 0.002419667
hierarchical model 0.002410316
newswire data 0.002407305
ing data 0.002404281
source phrase 0.002385093
data quality 0.002371677
data selection 0.002363915
mixture model 0.002329839
model parameters 0.002311782
data types 0.00228744
target phrase 0.0022682120000000004
log phrase 0.002260529
guage model 0.002255481
data origin 0.002253354
model adap 0.002241191
conversational data 0.002233017
data track 0.002231906
data all 0.002231244
tic data 0.002229876
data collections 0.002229876
unconstrained data 0.002229876
data consortium 0.002229876
data release 0.002229876
phrase pairs 0.002149541
word alignment 0.002126344
phrase rule 0.00207778
phrase pair 0.002012095
same word 0.001974151
weighted phrase 0.001972888
relevant phrase 0.001962984
model 0.00195239
phrase transla 0.001945127
unique phrase 0.00194092
sentence weight 0.001934491
sentence features 0.0019200749999999998
corpus weight 0.001883601
sentence weights 0.0018108679999999998
smt system 0.00180969
optimization method 0.0017784839999999999
corpus weights 0.0017599780000000001
