translation model 0.0039413
translation rule 0.00304704
target translation 0.0029161300000000003
context word 0.0028664800000000002
language model 0.002738431
space model 0.002684244
word sense 0.002660067
translation system 0.002658315
translation rules 0.0025977969999999998
bilingual word 0.002487874
ibm model 0.002431542
translation method 0.002409439
translation probability 0.002407197
translation performance 0.002333847
machine translation 0.002331614
tion model 0.002309408
translation pairs 0.002300993
translation probabilities 0.002279503
translation lexicon 0.00225915
word pair 0.00225843
word clustering 0.002252115
source words 0.0022128399999999998
particular translation 0.002205654
translation mod 0.002203138
context words 0.0022019500000000003
translation selection 0.002201814
word alignment 0.002199902
target words 0.00219905
translation modeling 0.002198267
chinese word 0.002188262
translation units 0.0021755439999999997
translation perfor 0.002172432
translation proba 0.00216116
translation unit 0.0021607609999999998
wrong translation 0.00216064
translation accu 0.002159153
source rule 0.0021224
training data 0.0021087099999999998
wrong word 0.00210809
word than小溪 0.002106998
source feature 0.00202295
context features 0.0020181599999999997
target features 0.00201526
model 0.00201402
context feature 0.00201206
target feature 0.0020091600000000003
source context 0.00199439
source target 0.0019914900000000003
target context 0.0019806
translation 0.00192728
phrase translations 0.0018970750000000002
similarity feature 0.001888162
different words 0.001879746
training corpus 0.001862531
phrase pair 0.0017322700000000002
source language 0.0017270509999999998
target language 0.0017132610000000002
target vector 0.001699799
different source 0.001672186
different context 0.001661296
initial phrase 0.001656969
individual phrase 0.0016556210000000001
sense similarity 0.001653189
particular phrase 0.0016269440000000002
language models 0.001617091
parallel training 0.001609992
large data 0.0015730990000000001
hierarchical rule 0.00156711
data task 0.0015144709999999999
rule pair 0.00150346
rule pairs 0.001493473
stop words 0.00148365
rule table 0.001482931
bilingual similarity 0.001480996
other similarity 0.001477679
full training 0.001476268
following features 0.001473839
nist training 0.001457297
synonymous words 0.00144308
feature vectors 0.001429297
feature weight 0.001417533
distance similarity 0.001409558
source side 0.001401868
context vectors 0.001400737
bilingual sense 0.001398481
target vectors 0.001397837
data sets 0.001397582
language space 0.001394635
different language 0.001393957
simple features 0.0013923219999999999
similarity function 0.001392278
feature weighting 0.001389956
target side 0.001388078
vector space 0.0013811729999999999
rate training 0.0013766339999999998
new similarity 0.001370113
data nist 0.001369907
meeting rule 0.001364917
contextual features 0.001359627
scfg rule 0.001357328
