phrase translation 0.00656392
phrase pair 0.004730394
phrase table 0.004656465
phrase pairs 0.004463445
source phrase 0.00443993
new phrase 0.00442099
phrase length 0.004395509
translation model 0.00428616
long phrase 0.004215758
phrase number 0.004206122
phrase transla 0.004176121
phrase coverage 0.00417431
unseen phrase 0.004157921
upper phrase 0.004124106
whole phrase 0.0041100600000000004
chical phrase 0.004088128
phrase 0.0038092
translation weights 0.003231969
machine translation 0.003212939
translation quality 0.0031712900000000002
translation options 0.003038513
translation template 0.003033764
chine translation 0.003030712
ble translation 0.003029587
translation option 0.003026919
word alignment 0.00299001
translation 0.00275472
source phrases 0.00249738
bilingual phrases 0.002363341
similar phrases 0.002328496
language model 0.002306534
source word 0.00230406
long phrases 0.002273208
word alignments 0.002246945
unseen phrases 0.0022153710000000003
chinese word 0.002214608
short phrases 0.0021911640000000002
training corpus 0.002142476
tal phrases 0.0021396410000000003
word reordering 0.002136781
word segmentation 0.002081088
correct word 0.0019665810000000002
alignment information 0.001957244
word substitution 0.001950454
record word 0.0019457559999999999
word substitu 0.0019457559999999999
phrases 0.00186665
pbsmt model 0.001866487
test data 0.00185535
alignment matrix 0.001841217
template model 0.001810484
training cor 0.001656476
training script 0.001645719
alignment template 0.001595724
gual data 0.0015862250000000001
data sparseness 0.0015760190000000001
model 0.00153144
serious data 0.0015214
different words 0.001415679
training 0.00137323
tion pair 0.001354653
task table 0.001321769
alignment 0.00131668
same length 0.001309412
test set 0.001300942
sentence pairs 0.001289359
bilingual corpus 0.001265937
source sentence 0.001265844
large corpus 0.001265568
new feature 0.0012649929999999998
small corpus 0.001184198
ing corpus 0.001180061
chinese sentence 0.001176392
grammar table 0.0011702050000000001
search algorithm 0.001166051
language modeling 0.001158201
words classes 0.001157038
feature weights 0.001130452
pruning parameter 0.001122735
same pos 0.001108076
matching threshold 0.0011019979999999999
development set 0.00109244
bleu score 0.001073905
matching method 0.001049862
sri language 0.001049501
gigaword corpus 0.001044722
trainig corpus 0.001041615
pbsmt models 0.001039456
matching methods 0.0010381359999999998
single source 0.001029124
baseline system 0.001024554
long distance 0.001013798
matching similarity 0.001013677
evaluation test 0.001010866
source side 0.00100644
position set 9.99213E-4
source text 9.96644E-4
long term 9.78257E-4
other hand 9.5882E-4
bleu scores 9.583230000000001E-4
