translation model 0.00436198
word alignments 0.0034947399999999997
english word 0.0032398360000000003
language model 0.003159737
translation system 0.003076159
phrasal translation 0.002795132
word order 0.0027809560000000002
statistical translation 0.002735576
czech word 0.002718386
machine translation 0.002700546
lexical translation 0.0026785050000000003
ibm model 0.002675374
model trans 0.002671936
model parameters 0.002667356
translation process 0.002613469
translation quality 0.002609951
transformation model 0.002565943
extra word 0.002565839
lation model 0.002552641
english feature 0.002490416
translation prob 0.0024780320000000002
entire translation 0.0024585600000000003
reference translation 0.002453676
model 0.00223427
tory alignment 0.002229633
translation 0.00212771
different features 0.002043972
english dependency 0.002037796
english tree 0.002022906
czech feature 0.001968966
alignment 0.00193624
possible alignments 0.001921059
feature values 0.001908154
several features 0.001895486
feature description 0.001863755
english phrase 0.0018530600000000001
feature com 0.001849357
feature φei 0.00180686
feature φfi 0.001803145
english sentence 0.001746035
english data 0.0017440630000000001
analytical features 0.001683908
dependency structure 0.0016686600000000002
dependency node 0.001658465
statistical tree 0.001647996
dependency nodes 0.001619764
function words 0.001608611
english node 0.001586221
dependency structures 0.0015596400000000002
english nodes 0.0015475200000000001
other words 0.001543104
czech dependency 0.001516346
feature 0.00150764
dependency parser 0.001505063
english structures 0.0014873960000000002
system models 0.0014813019999999999
same probability 0.001464908
same training 0.0014319699999999999
system performance 0.001430894
features 0.00138785
english rules 0.001385355
phrase structures 0.0013749040000000001
tree transformation 0.001371803
foreign language 0.00136753
english part 0.0013619770000000001
dependency treebank 0.001356029
english parse 0.0013524960000000001
togrammatical dependency 0.0013486980000000002
prague dependency 0.0013486980000000002
english string 0.001346909
pennsylvania tree 0.00133388
czech phrase 0.00133161
english side 0.001330074
corresponding english 0.001324681
partial phrase 0.001314111
entire english 0.001313626
eign words 0.001307027
system errors 0.001291502
system output 0.001283413
parallel corpus 0.001281089
sentence pairs 0.001273223
first work 0.001272973
system overview 0.001261316
own system 0.00124432
benchmark system 0.00124354
alignments 0.00123768
czech sentence 0.001224585
czech data 0.001222613
inflected language 0.001218884
statistical machine 0.001180702
phrase span 0.001168961
probability mass 0.0011678600000000002
respective phrase 0.001164882
head position 0.001154794
bleu score 0.001138604
fiscal figure 0.0011334370000000002
training conditions 0.0011240080000000001
syntactic structure 0.001096194
following information 0.001087644
possible values 0.001083893
