topic word 0.00455241
word topic 0.00455241
topic model 0.00432237
topic words 0.00391202
topic distribution 0.00374451
word translation 0.0036238399999999997
topic cache 0.00353978
topic language 0.00341976
translation model 0.0033937999999999998
same topic 0.003290852
topic modeling 0.0032258169999999997
topic distributions 0.003138227
relevant topic 0.003055673
economic topic 0.003041845
topic dis 0.00304042
topic caches 0.003022754
topic distri 0.003003717
major topic 0.003001709
topic mod 0.002996917
word distribution 0.00282604
topic 0.00273544
bilingual translation 0.002678819
translation models 0.002620001
bilingual phrase 0.002445589
machine translation 0.0024115
word pair 0.002385409
word alignment 0.002373423
translation modeling 0.0022972469999999997
translation results 0.002284313
english translation 0.002271254
language model 0.00227125
translation pairs 0.002245339
phrase candidate 0.002234336
target phrase 0.0022198310000000002
word distributions 0.002219757
translation task 0.002194915
translation process 0.002173211
translation probabilities 0.002171321
translation candidates 0.002155525
specific word 0.002145745
phrase pair 0.002142079
specific translation 0.002135645
phrase alignment 0.002130093
consistent translation 0.0021236939999999998
translation systems 0.0021144799999999997
translation hypotheses 0.002112351
possible translation 0.002102431
translation quality 0.00209916
word penalty 0.0020863039999999998
translation hypothesis 0.002083536
jth word 0.002067779
phrase table 0.002041324
lda model 0.002033801
local model 0.00203302
source phrase 0.002030491
phrase pairs 0.002012109
ing model 0.0020025340000000003
related phrase 0.001955505
phrase transla 0.001912722
relevant phrase 0.0018938730000000001
training data 0.0018881219999999999
probabilistic model 0.001887552
phrase level 0.0018715770000000001
guage model 0.0018687460000000001
gual phrase 0.0018573910000000001
phrase segmentation 0.001837066
lingual phrase 0.001834363
long phrase 0.0018336910000000001
likely phrase 0.001827455
merging phrase 0.00182317
target words 0.001822771
translation 0.00180687
training document 0.001787111
training method 0.0016209689999999999
cache feature 0.001616011
words pairs 0.001615049
bilingual document 0.00160805
model 0.00158693
training parallel 0.001573734
phrase 0.00157364
bilingual sentence 0.0015332050000000002
test data 0.001525116
language models 0.001497451
document corpus 0.0014243099999999998
test document 0.001424105
feature function 0.001421912
smt system 0.001401234
different test 0.001392144
target document 0.0013822919999999998
new feature 0.001362919
whole training 0.001333775
target language 0.001330511
discriminative training 0.001307409
target candidate 0.001306887
document pair 0.0013045399999999999
same document 0.001291513
new document 0.001287349
domain language 0.0012825290000000001
smt research 0.001276054
cache table 0.001272024
