tree rule 0.0045807
translation rules 0.00418458
translation forest 0.00396765
tree nodes 0.0038499600000000004
source tree 0.003466129
target tree 0.0034454710000000003
tree models 0.003251636
frontier tree 0.003208959
tree pairs 0.0031864750000000002
tree probabilities 0.0031535870000000002
synchronous tree 0.0031446490000000002
tree pair 0.0031386110000000004
tree root 0.0029771520000000003
tree substitution 0.00294925
tier tree 0.0029104760000000004
tree leaves 0.00290754
forest pruning 0.00274868
machine translation 0.002704763
translation quality 0.002695918
noisy translation 0.002671316
translation qual 0.0026526709999999997
superior translation 0.0026519269999999997
source node 0.002558529
target node 0.002537871
rule probabilities 0.002531027
rule extraction 0.002522206
minimal rule 0.0025187409999999997
source forest 0.002485609
target forest 0.002464951
parse forest 0.002447983
new rule 0.0024352619999999997
rules bleu 0.002412271
minimal rules 0.0023777110000000002
rule table 0.002353071
second rule 0.00234799
translation 0.00234654
stsg rule 0.002337268
rule coverage 0.002303012
frontier node 0.0023013590000000002
third rule 0.0022972039999999997
reasonable rule 0.002292106
refinement rule 0.0022848549999999997
estimating rule 0.0022843019999999998
initial rules 0.002233138
forest pairs 0.002205955
abstract rules 0.002203044
stsg rules 0.002196238
unary rules 0.002194037
minimum rules 0.002188729
useful rules 0.002181962
forest pair 0.002158091
imal rules 0.002149776
mal rules 0.002144244
composed rules 0.002143472
production rules 0.002143472
node count 0.002051491
maximal node 0.002043596
packed forest 0.0020280890000000003
different pruning 0.002017174
example forest 0.002016291
consistent node 0.002009757
imal node 0.002005766
distinguished node 0.002002824
node attributes 0.002001213
language model 0.001967487
lation forest 0.0019527490000000002
pack forest 0.0019265950000000001
model figure 0.0018929569999999998
structure trees 0.0018846140000000002
frontier nodes 0.001855659
rules 0.00183804
phrase structure 0.001811041
model score 0.0017190769999999998
shaded nodes 0.001700771
target trees 0.001694095
source parse 0.001691372
forest 0.00162111
dependency trees 0.00161393
training data 0.0016007679999999998
cube pruning 0.001580362
original model 0.001575211
guage model 0.001572402
pruning method 0.001570037
model outper 0.00155876
tween nodes 0.001556806
tail nodes 0.001556518
many phrase 0.001546852
pruning threshold 0.001528962
extraction algorithm 0.001517286
chinese sentence 0.00149175
source frontier 0.0014718280000000001
pruning thresholds 0.001468657
english sentence 0.0014598229999999998
frontier trees 0.0014575830000000001
source words 0.0014536100000000001
target frontier 0.0014511699999999999
word alignments 0.001450265
pruning thresh 0.001434374
target words 0.001432952
test set 0.001418039
