alignment model 0.005175320000000001
word alignment 0.00482709
translation model 0.00444075
alignment models 0.0038337600000000003
word alignments 0.0034976399999999998
alignment approach 0.003444906
alignment probability 0.003400356
corpus model 0.0032336350000000003
same model 0.003211075
ibm model 0.003186998
alignment probabilities 0.003174825
language word 0.0031595879999999996
current alignment 0.003153232
model parameters 0.0031407880000000003
statistical alignment 0.003105299
translation models 0.00309919
alignment problem 0.003094665
other model 0.003061831
alignment approaches 0.0030491520000000003
model parameter 0.0030356230000000003
alignment mod 0.003033988
alignment systems 0.002985022
direct alignment 0.002975936
alignment proba 0.0029743490000000003
further model 0.002961103
empty alignment 0.002959097
statistical model 0.002955519
previous model 0.002954885
tical alignment 0.002954215
alignment strategies 0.002952578
partial alignment 0.002951164
hmm model 0.002933713
target word 0.002888506
source word 0.002887255
fixed model 0.002869158
model scaling 0.002859639
transition model 0.002852667
model parame 0.0028374290000000003
extended model 0.002825169
model scal 0.002801379
word align 0.0027721819999999998
alignment 0.00266255
word order 0.002648425
word pairs 0.002605606
ibm translation 0.002602208
various word 0.002597804
traditional word 0.002544162
model 0.00251277
french word 0.0024805
word positions 0.002458916
langauge word 0.002457433
word eaj 0.002453995
statistical translation 0.002370729
machine translation 0.002353765
translation mod 0.002299418
translation process 0.002279576
translation direc 0.002247222
translation directions 0.002238072
translation probabil 0.002221899
training data 0.00216858
function words 0.002124397
ibm alignments 0.002007328
translation 0.00192798
heuristic models 0.001925787
feature function 0.001922255
language sentence 0.001898213
source words 0.001868625
ibm models 0.001845438
new features 0.001823322
possible alignments 0.00177363
linear models 0.0017723069999999999
training corpus 0.001764205
heuristic function 0.001733064
target language 0.001719014
source language 0.0017177629999999998
other words 0.001694971
training method 0.001693475
viterbi alignments 0.001675553
worse alignments 0.001646097
annotated alignments 0.001638704
ble alignments 0.001632594
target sentence 0.001627131
source sentence 0.0016258800000000001
statistical models 0.0016139589999999999
hmm models 0.001592153
corresponding words 0.001542053
tion models 0.001541948
cal models 0.001500674
lation models 0.001491771
probabilistic models 0.0014892759999999999
sentence pair 0.001460284
models brown 0.001459613
features 0.00145928
bilingual sentence 0.0014360689999999999
guistic data 0.001414342
function gain 0.0013939870000000002
training scheme 0.001388455
feature functions 0.001367982
large set 0.001358278
different gain 0.001352528
