translation model 0.0045495399999999995
source phrase 0.00395155
target phrase 0.0039341
source translation 0.00355981
phrase pair 0.003513195
phrase pairs 0.003508088
phrase table 0.003505311
other phrase 0.003485645
phrase representation 0.003458826
phrase length 0.003454654
output phrase 0.003447967
large phrase 0.003403344
many phrase 0.003399132
input phrase 0.00336745
english phrase 0.003313004
ing phrase 0.003305938
phrase scores 0.003290412
phrase representations 0.003283281
translation probability 0.003276406
maximum phrase 0.003253998
phrase level 0.003249815
french phrase 0.003209065
unique phrase 0.003202803
actual phrase 0.003193266
scoring phrase 0.003188165
translation task 0.003155415
machine translation 0.0030761259999999998
translation system 0.0030665889999999998
translation performance 0.0030621119999999996
translation probabilities 0.003052993
translation hypotheses 0.002957238
phrase 0.00291186
overall translation 0.002815416
translation probabil 0.002799467
direct translation 0.002799181
uous translation 0.0027987819999999997
ence translation 0.0027965779999999997
translation perfor 0.0027965779999999997
language model 0.002770583
new model 0.002679341
similar model 0.002611832
model rnn 0.002567558
translation 0.00252012
network model 0.002504917
model parameters 0.002491283
word probability 0.002488056
decoder model 0.002474366
model log 0.002455019
novel model 0.002391831
trained model 0.002357305
linear model 0.002356571
lation model 0.002339354
systems model 0.002321628
model architecture 0.002317104
cslm model 0.0023058090000000002
word representation 0.002278736
input word 0.00218736
word penalty 0.002107899
word representations 0.002103191
word counts 0.00207696
same words 0.002073696
word level 0.002069725
model 0.00202942
phrases source 0.0020267690000000003
source phrases 0.0020267690000000003
similar words 0.002022092
target phrases 0.002009319
training set 0.001961101
context words 0.001903045
source sentence 0.001875892
training sta 0.0017860699999999998
target language 0.001763403
french words 0.001736885
unknown words 0.001729017
target corpus 0.00169622
such source 0.001656786
neural language 0.001638096
phrases table 0.0015805300000000001
sequence data 0.001574636
bleu score 0.001545574
target sentences 0.001522323
training 0.00150786
space vector 0.0015051230000000001
source context 0.001503055
source sequence 0.0014979659999999999
target sequence 0.0014805159999999999
words 0.00143968
data selection 0.001439101
frequent source 0.001434472
penalty source 0.001415819
space language 0.001408461
vector representation 0.001384791
frequent phrases 0.001381861
length vector 0.001380619
target side 0.001375323
neural network 0.00137243
rare source 0.001345143
linguistic phrases 0.0013398280000000002
additional feature 0.001336732
language models 0.001336432
