source word 0.0040689
translation model 0.0040655199999999996
target word 0.00380521
word alignment 0.0036368670000000002
translation rule 0.0033651
head word 0.0033574860000000002
second word 0.003328376
word trigger 0.003224788
parent word 0.003199293
ond word 0.003145712
language model 0.0030223209999999997
translation rules 0.003004459
translation probability 0.002831552
dependency models 0.002813421
translation system 0.002809277
dependency feature 0.002785865
structure dependency 0.002761318
dependency structure 0.002761318
dependency tree 0.002717657
lexical translation 0.00256295
dependency information 0.00252816
hierarchical translation 0.002510126
translation performance 0.002505256
entropy model 0.0024717209999999996
ibm model 0.0024647669999999997
translation probabilities 0.0024351800000000003
smt model 0.0024290249999999996
dependency constraint 0.0024001120000000003
translation quality 0.002397389
dependency parser 0.002394611
gram model 0.00239421
machine translation 0.002388561
dependency relation 0.0023840140000000003
translation task 0.002383361
dependency edge 0.002350687
hpb translation 0.002347666
monolingual dependency 0.002344437
formed dependency 0.002323515
translation prob 0.002321544
translation result 0.002315122
dependency struc 0.002314014
dependency rela 0.002285733
dependency relationship 0.0022824110000000002
dependency restriction 0.002278032
source sentence 0.00224682
source target 0.00215147
system rule 0.002103437
translation 0.00203547
model 0.00203005
dependency 0.00200182
target sentence 0.00198313
hierarchical rule 0.001804286
maximum rule 0.001798102
rule table 0.001787807
training corpus 0.00177591
rule length 0.0017081610000000001
rule extraction 0.0016862350000000001
rule selection 0.0016756470000000002
source side 0.001643949
rule entries 0.0016241930000000001
hierarchal rule 0.001623191
corresponding source 0.001615354
rule extrac 0.001610481
erarchical rule 0.001608478
inal rule 0.001606747
context words 0.001569893
english words 0.001569448
first training 0.001548969
additional language 0.001522684
source sen 0.001517954
context language 0.001508154
sentence pairs 0.0014793200000000001
large corpus 0.001479127
grammar rules 0.001459656
hierarchical rules 0.001443645
chinese words 0.001443399
language modeling 0.0014263840000000002
target side 0.001380259
rate training 0.001371948
semantic structure 0.0013642810000000002
language trigger 0.0013557390000000001
tree structures 0.001354529
training corpora 0.001351779
unaligned words 0.001340027
parallel training 0.001334091
glish words 0.001332461
oov words 0.001331409
plain sentence 0.001323539
distance reordering 0.001323447
baseline system 0.001317978
pendency language 0.001305617
sri language 0.001275738
new feature 0.001272246
feature score 0.001266396
ing corpus 0.001262966
new system 0.001262008
bad rules 0.001254297
chical rules 0.001248482
data sparseness 0.001218957
source 0.00120758
