source word 0.002805885
sequence model 0.002450565
word sequence 0.002376515
word error 0.002315668
transformation model 0.002294303
probabilistic model 0.002193015
global model 0.002175845
tion model 0.002152269
unigram model 0.002147172
verbatim word 0.002146209
mation model 0.0021326139999999997
channel model 0.002121077
training data 0.002104347
minimum word 0.00208282
source words 0.0020706450000000003
word loss 0.002061733
target words 0.002052181
word risk 0.002046793
word patient 0.002046793
word processor 0.002046793
source target 0.0019372859999999999
model 0.00190206
function words 0.0015857900000000001
source sequence 0.00152638
target sequence 0.001507916
test data 0.001496197
number words 0.001461347
text document 0.001457203
target pair 0.00143874
multiple words 0.0013727960000000001
training documents 0.001362939
source position 0.001336559
nized words 0.001311585
ordinary words 0.001311585
alignment step 0.0013074699999999998
new document 0.001288517
source sequences 0.001264942
corresponding source 0.0012602849999999999
target tokens 0.001258247
distance alignment 0.001255847
entire source 0.001241855
factored source 0.001213291
empty target 0.001191855
document length 0.001186639
sentence boundary 0.0011819999999999999
alignment scheme 0.0011755749999999999
different text 0.001160235
document manual 0.001132936
final document 0.0011162770000000002
ical document 0.001093157
words 0.00109277
draft document 0.001089612
standard language 0.001082572
compliant document 0.001075055
short phrase 0.001071522
language modeling 0.001057758
joint probability 0.001023319
language processing 0.00100844
text transformation 9.93463E-4
source 9.77875E-4
acoustic models 9.70667E-4
formation models 9.67747E-4
rule set 9.670099999999999E-4
target 9.59411E-4
different learning 9.57524E-4
baseline rule 9.56316E-4
natural language 9.562399999999999E-4
training 9.53277E-4
necessary information 9.421799999999999E-4
text segmentation 9.27511E-4
posterior probability 9.24813E-4
alignment 9.23906E-4
sentence 9.08815E-4
sequence position 9.071890000000001E-4
rior probability 8.783549999999999E-4
different techniques 8.674889999999999E-4
text seg 8.58607E-4
example documents 8.58045E-4
document 8.55983E-4
maximization algorithm 8.44079E-4
probabilistic sequence 8.3946E-4
test set 8.37331E-4
small sequence 8.32314E-4
human table 8.27633E-4
initial rule 8.24786E-4
text templates 8.23504E-4
token sequence 8.19881E-4
single rule 8.18165E-4
error rate 8.054010000000001E-4
replacement sequence 8.04577E-4
transformation method 7.968420000000001E-4
experimental evaluation 7.96462E-4
rule induction 7.94667E-4
decision rule 7.86341E-4
recognition task 7.81886E-4
empty sequence 7.80949E-4
different amounts 7.800519999999999E-4
small set 7.76013E-4
transformation step 7.75807E-4
transformation modeling 7.71444E-4
