target word 0.00479471
word alignment 0.00422866
additional word 0.0036309759999999997
recall word 0.00345854
word link 0.003456952
general word 0.003428759
target target 0.00341454
word links 0.003366106
underlying word 0.0033402199999999996
english translation 0.003098211
target source 0.00303606
source target 0.00303606
machine translation 0.003027914
translation performance 0.002935007
target phrase 0.0028746500000000003
translation task 0.002822271
language model 0.002821208
translation systems 0.002750749
translation quality 0.002750197
source source 0.00265758
target words 0.00253929
target language 0.002501498
translation 0.00245825
model parameters 0.002376128
unigram model 0.002371616
joint model 0.002288175
target figure 0.0022340470000000003
source words 0.00216081
first target 0.002143393
target phrases 0.002123462
source language 0.002123018
maximum target 0.0021229820000000003
block alignment 0.002081883
model 0.00202698
training data 0.0020037049999999997
target interval 0.002003157
minimum target 0.001992456
target positions 0.001989217
target intervals 0.001975852
target index 0.001970405
contiguous target 0.00196603
target clump 0.0019637640000000002
sentence block 0.001864285
source figure 0.001855567
alignment function 0.0017697390000000002
source phrases 0.001744982
target 0.00170727
similar phrase 0.001689726
ing training 0.001633552
source interval 0.0016246770000000002
source positions 0.001610737
source intervals 0.0015973720000000001
intervals source 0.0015973720000000001
baseline block 0.001590118
contiguous source 0.00158755
projects source 0.0015812600000000001
denote source 0.0015812600000000001
recall alignment 0.0015123200000000002
alignment link 0.001510732
alignment relations 0.0014431070000000001
alignment func 0.0014411580000000001
viterbi training 0.001435016
alignment links 0.001419886
union alignment 0.001393909
alignment templates 0.001393909
block sequence 0.00138378
source 0.00132879
block learning 0.0013239319999999999
block link 0.001310175
unigram block 0.001285299
block unigram 0.001285299
block generation 0.001258658
precision block 0.001249511
novel block 0.001223169
block extension 0.0012201199999999999
block links 0.001219329
extend block 0.001206693
block decod 0.001200296
extended block 0.001196487
block exten 0.001193625
phrase 0.00116738
alignment 0.00114122
training 0.00114085
similar system 0.001100289
trigram language 0.001095348
learning algorithm 0.0010796920000000002
same number 0.001064544
system performance 0.0010547
statistical machine 0.001021336
generation algorithm 0.001014418
possible blocks 0.001002926
extension algorithm 9.7588E-4
sequence figure 9.69894E-4
algorithm empirically 9.51618E-4
sion algorithm 9.51618E-4
tension algorithm 9.51618E-4
block 9.40663E-4
other parameters 9.37869E-4
sentence 9.23622E-4
extension approach 8.75889E-4
