translation model 0.00483144
reordering model 0.00389385
different model 0.003354239
language model 0.003324634
ing model 0.003147174
orientation model 0.0030063010000000003
model scores 0.002998208
baseline model 0.002983472
model score 0.002979412
tion model 0.002934116
translation probability 0.0029078890000000003
dominance model 0.002907441
graph model 0.0028910990000000003
probabilistic model 0.002884062
model this 0.002883658
model pdomo 0.00287871
separate model 0.002871162
translation rules 0.0028680430000000002
model newswire 0.002867041
model spans 0.002857329
bigram model 0.002853407
unigram model 0.002846681
centered model 0.002840739
nance model 0.002840739
arate model 0.002840739
translation system 0.002813475
translation task 0.00269185
other translation 0.00264729
machine translation 0.002646709
model 0.00260698
lexical translation 0.002599746
multiple translation 0.002581517
translation probabilities 0.002527122
translation units 0.002526585
translation unit 0.002498207
translation mod 0.0024869840000000002
ple translation 0.0024859360000000002
actual translation 0.002467319
tiple translation 0.002463101
mal translation 0.002458294
source word 0.0023734999999999997
word alignment 0.002241279
translation 0.00222446
target word 0.002223768
different reordering 0.002034129
same reordering 0.001937126
source phrase 0.001924777
new features 0.0018728459999999998
source sentence 0.001872528
function word 0.001844934
syntactic features 0.001841573
tic features 0.0017986959999999998
compound features 0.0017754659999999998
target phrase 0.0017750449999999998
porate features 0.001739538
elementary features 0.0017366439999999999
tual features 0.0017342619999999999
provenance features 0.0017342619999999999
target sentence 0.001722796
next word 0.001708123
important reordering 0.0017071550000000001
local reordering 0.001692393
reordering process 0.0016788250000000001
previous word 0.001661054
word level 0.001651104
source sentences 0.001648571
target language 0.001639282
monotone reordering 0.0016337970000000002
training data 0.001600321
corresponding word 0.001595084
usa reordering 0.001560802
source order 0.0015604949999999999
word class 0.001559342
internal word 0.001558266
actual word 0.001544999
uniform reordering 0.001543452
inverse reordering 0.001531869
actual reordering 0.001529729
reordering pattern 0.0015290500000000001
cast reordering 0.0015288340000000002
reordering patterns 0.001524509
plex reordering 0.0015234060000000002
reordering challenge 0.001523239
calized reordering 0.001520974
features 0.00149929
hierarchical phrase 0.001449194
language models 0.001445122
target order 0.001410763
additional feature 0.0013978469999999998
feature scores 0.001396098
target side 0.001391607
feature score 0.001377302
corresponding source 0.001364304
source index 0.00132625
phrase pair 0.001325707
english words 0.001317968
ing algorithm 0.001300003
reordering 0.00128687
target string 0.0012834119999999998
feature weights 0.0012795859999999999
