word model 0.0035896400000000003
word alignment 0.00293545
probability model 0.0028751790000000003
alignment models 0.00272781
nondeficient model 0.002531559
distortion model 0.0025180610000000003
original model 0.002517656
model variants 0.0025142790000000003
based model 0.0024553720000000004
respective model 0.002419136
probabilistic model 0.002412828
regular model 0.0023838630000000004
guage model 0.0023830550000000002
base model 0.0023786450000000004
former model 0.0023658620000000003
model iter 0.0023583090000000003
parametric model 0.00235737
der model 0.00235503
model degree 0.00235503
first word 0.002181976
model 0.00209169
english word 0.0020471120000000002
such models 0.002027857
empty word 0.001932013
same alignment 0.0019297519999999999
foreign word 0.001875612
ibm models 0.001871849
models ibm 0.001871849
target word 0.001850709
viterbi alignment 0.001846414
word class 0.001823219
fertility models 0.001795132
word classes 0.001783714
current alignment 0.001766798
alignment accuracy 0.0017652549999999999
glish word 0.001762906
likely alignment 0.001756311
nondeficient models 0.001730179
alignment accuracies 0.001726973
distortion models 0.001716681
alignment literature 0.001706469
alignment computation 0.001705696
linear models 0.001671419
deficient models 0.0016361330000000001
metric models 0.001631797
complex models 0.00162923
training data 0.0015852969999999998
conditional models 0.001567659
dard models 0.001554219
alignment 0.0014375
translation performance 0.001420392
same problem 0.001335818
data structures 0.001325741
models 0.00129031
based translation 0.001280659
optimization problem 0.001275875
following words 0.001275259
probability distribution 0.0012692979999999999
following probability 0.001260514
many words 0.001248092
translation probabilities 0.001245893
empty words 0.001232297
maximization problem 0.001226485
refined data 0.001211092
distortion probability 0.0012098599999999999
translation systems 0.001208434
problem max 0.0012044080000000001
english sentence 0.001202651
corpus probability 0.001202302
proper algorithm 0.00118937
source words 0.001184412
foreign words 0.001175896
likelihood problem 0.001169755
objective function 0.0011542969999999999
terms problem 0.001153973
function value 0.001136869
first order 0.001135448
cient problem 0.001125495
hard problem 0.001116475
problem och 0.001107984
unaligned words 0.001106601
strategy words 0.001094819
sentence pair 0.001093911
probability distri 0.00108625
base probability 0.0010704439999999998
rare words 0.001063768
probability normalization 0.001055012
start probability 0.001054122
probability mass 0.001053877
normalization constraints 0.001051745
malization constraints 0.001046133
first iteration 0.001044109
above function 0.0010351520000000001
sentence pairs 0.001033694
foreign sentence 0.0010311510000000001
constraint set 0.001024922
first task 0.001023808
other number 0.001016561
training scheme 9.87836E-4
rate training 9.852469999999999E-4
