word alignment 0.0037654900000000002
alignments alignment 0.0032434100000000004
word alignments 0.00295796
alignment data 0.002921951
alignment error 0.002525737
source word 0.002500196
alignment task 0.002485733
alignment matrix 0.002475489
translation models 0.002316479
suitable alignment 0.00228915
alignment matrices 0.002288004
statistical model 0.002146434
correct model 0.002115712
null word 0.002101033
model size 0.0020866540000000003
alignment 0.00202547
probabilistic model 0.002025131
translation matrix 0.002021279
cal model 0.002016929
translation table 0.001997664
model selection 0.001989503
mixture model 0.0019857580000000002
model complexity 0.001983876
machine translation 0.001943176
translation mod 0.0018660550000000001
chine translation 0.0018369810000000001
nal translation 0.0018339510000000001
english words 0.001814797
target words 0.0017365640000000001
model 0.00172144
proper alignments 0.001652719
other words 0.0016136280000000002
null alignments 0.0015789530000000001
translation 0.00157126
cepts alignments 0.0015579600000000002
test data 0.001533076
factorisation alignments 0.00152845
sure alignments 0.001528365
source sentence 0.001511048
reference alignments 0.001510962
probable alignments 0.001490573
able alignments 0.001483008
target sentence 0.0014766760000000001
words figure 0.001415355
null words 0.0013717730000000002
words cepts 0.0013507800000000002
french words 0.00131131
alignments 0.00121794
reference data 0.001189503
trial data 0.001181631
ibm models 0.001146595
test set 0.001088907
lation models 0.001077783
sentence level 0.001065602
first step 0.001047235
entire source 0.00103834
target sentences 0.001036821
plsa models 0.001034229
entire corpus 0.001020344
standard approach 0.001018671
complete models 0.001014629
words 0.00101076
different set 9.763790000000001E-4
many cases 9.687579999999999E-4
joint probability 9.552930000000001E-4
conditional probability 9.44566E-4
other cepts 9.42888E-4
possible number 9.38314E-4
following problem 9.36349E-4
first cept 9.32192E-4
simple example 9.22229E-4
additional matrix 9.179730000000001E-4
other side 9.08322E-4
novel algorithm 9.027390000000001E-4
large values 8.98272E-4
important problem 8.90196E-4
first factorise 8.779980000000001E-4
other hand 8.73134E-4
different cepts 8.64087E-4
shared task 8.60332E-4
error rate 8.53347E-4
different languages 8.42615E-4
different weights 8.40644E-4
tion matrix 8.221120000000001E-4
proper align 8.19632E-4
english 8.04037E-4
uniform distribution 7.97497E-4
standard nmf 7.97184E-4
low value 7.89974E-4
count matrix 7.846940000000001E-4
association measure 7.75312E-4
linguistic annotation 7.73066E-4
selection problem 7.6853E-4
ficult problem 7.668670000000001E-4
performance measures 7.61741E-4
matrix factorisation 7.60529E-4
source 7.60176E-4
factorisation standard 7.5432E-4
second step 7.52144E-4
sentence 7.50872E-4
