translation word 0.005122359999999999
target word 0.00419685
source word 0.00413135
translation models 0.00356902
head word 0.003420487
translation system 0.003417016
transducer word 0.003121726
translation accuracy 0.00308551
speech translation 0.003072142
statistical translation 0.003069766
machine translation 0.00302575
dependent word 0.003006131
translation systems 0.00298165
chinese translation 0.002980675
translation methods 0.002965111
ing translation 0.002940457
text translation 0.002933517
word error 0.0029205729999999997
lation word 0.002914963
translation component 0.002850952
translation errors 0.002834576
translation mod 0.002832182
guage word 0.002825818
word labels 0.002825486
translation application 0.002823294
rect translation 0.0028147470000000003
nese translation 0.002810726
chine translation 0.002808391
translation applications 0.002806918
translation diver 0.002806918
target dependency 0.002603818
source language 0.00260078
translation 0.00255514
source dependency 0.002538318
transducer model 0.0024237160000000002
target words 0.002403882
model structure 0.002383865
model size 0.002379631
model transfer 0.002357447
transfer model 0.002357447
additional target 0.002258899
transduction model 0.002224695
automatic model 0.002200987
model complexity 0.002200193
additional source 0.002193399
model com 0.002180198
model complex 0.00214403
fixed model 0.002143091
model structures 0.002137294
model building 0.0021257430000000003
fer model 0.002122225
source tree 0.002086355
target relations 0.002079598
language models 0.00205053
target sequence 0.002035496
target string 0.002018729
right target 0.002012553
source sequence 0.001969996
second source 0.001963251
left target 0.001957519
source positions 0.001949802
right source 0.001947053
correct target 0.001940732
target sequences 0.001934496
target side 0.001924898
target representations 0.001905518
wrong target 0.001905173
target nodes 0.001896155
single source 0.001894582
left source 0.001892019
source sentences 0.001888991
target posi 0.001881285
source utterance 0.001877568
model 0.00186921
source sequences 0.001868996
source side 0.001859398
entire source 0.00184382
source lan 0.001843519
third source 0.001833472
source fragments 0.001832034
source frag 0.0018158979999999998
training data 0.001807711
training corpus 0.001771386
target 0.00162963
head words 0.001627519
transducer models 0.001568386
source 0.00156413
english head 0.001515589
transfer models 0.0015021169999999999
bilingual dependency 0.001497117
dependency representation 0.0014970909999999999
dependency tree 0.0014964129999999998
state head 0.001465193
generative models 0.001460247
dependency relations 0.001424156
transducer system 0.0014163819999999999
head transducer 0.001407773
training method 0.0014037350000000001
test data 0.001387145
language vocabulary 0.001376402
