word alignment 0.00558464
english word 0.00410852
word segmentation 0.003614533
word alignments 0.003570345
alignment models 0.003509664
previous word 0.003468568
alignment probability 0.003407418
traditional word 0.003351881
alignment set 0.00331811
hmm alignment 0.0031706240000000004
alignment approach 0.0031604800000000002
traditional alignment 0.0031454810000000003
alignment approaches 0.003142428
entity alignment 0.003120882
phrase alignment 0.0030783950000000003
language model 0.0030727099999999998
alignment position 0.003061039
alignment probabilities 0.0030096740000000004
alignment errors 0.00299085
final alignment 0.0029885800000000002
probable alignment 0.002952095
chinese translation 0.0028328900000000002
model parameters 0.002773758
hmm model 0.0027082139999999996
model hmm 0.0027082139999999996
alignment 0.00268912
entropy model 0.002653002
ibm model 0.0025950749999999996
maxent model 0.0025897679999999997
model parameter 0.002535336
channel model 0.002483184
translation probability 0.0024831280000000002
translation features 0.00246777
different translation 0.002441636
name translation 0.0022827200000000002
work translation 0.002274337
english words 0.00224945
translation score 0.002239575
model 0.00222671
machine translation 0.00219695
semantic translation 0.002170985
translation knowledge 0.002162146
translation characters 0.002159827
possible translation 0.002147719
translation list 0.002137759
translation scores 0.002134058
translation table 0.002109007
translation probabilities 0.002085384
translation cost 0.002064531
translation templates 0.002027629
translation strategies 0.002027023
translation template 0.00202694
english sentence 0.0019273650000000001
training data 0.00188924
chinese sentence 0.001782425
translation 0.00176483
english name 0.0017308900000000001
training corpus 0.001713027
data set 0.001619208
bilingual corpus 0.001609285
english person 0.0016030850000000002
english ner 0.001583545
language information 0.0015512669999999998
training set 0.001528012
english side 0.001521732
english string 0.0015208280000000001
english nes 0.001510367
data size 0.001481283
chinese character 0.001479846
english strings 0.0014778010000000002
surface english 0.0014725250000000001
english syllable 0.001469285
chinese characters 0.001463057
chinese person 0.001458145
chinese pinyin 0.001429465
new data 0.00139992
chinese side 0.001376792
target language 0.001374475
chinese nes 0.001365427
chinese help 0.001357912
large training 0.001324317
sparse data 0.001288214
language pairs 0.0012759339999999998
same sentence 0.00126789
segmented data 0.001255888
source language 0.001254758
seed data 0.001250192
data sparseness 0.001246982
parallel corpus 0.001244448
annotated training 0.001231053
english 0.001213
parameter training 0.001207648
segmentation error 0.001195066
bilingual knowledge 0.001192596
human language 0.001191665
feature function 0.0011898149999999999
ibm models 0.001188909
following sentence 0.001169572
target set 0.0011574649999999999
transliteration models 0.0011560799999999999
