translation model 0.00417727
training data 0.00405261
translation models 0.003672724
translation rules 0.0035694480000000002
similar translation 0.0035577710000000004
translation performance 0.0035392320000000002
training algorithm 0.003536768
translation results 0.003531526
translation task 0.003525795
translation weights 0.0034820800000000002
local training 0.003473571
other translation 0.0034722150000000003
training method 0.003432615
machine translation 0.003429393
translation examples 0.003423793
english translation 0.003406881
online translation 0.0033920580000000003
optimal translation 0.003335292
translation tasks 0.0033333520000000004
translation memory 0.0033194780000000003
translation framework 0.0033156590000000003
candidate translation 0.003307612
translation candidate 0.003307612
training corpus 0.003304397
good translation 0.00330209
translation result 0.0033004920000000004
ble translation 0.00329729
translation candidates 0.003284374
bad translation 0.00327457
translation exam 0.003271328
translation qualities 0.003269213
translation service 0.0032688720000000004
translation candi 0.003266966
translation candiates 0.003265843
translation services 0.003265843
translation consistency 0.003265843
pairwise translation 0.003265843
nbest translation 0.003265843
translation member 0.003265843
baseline training 0.003159039
global training 0.0031408670000000003
training step 0.003105762
training examples 0.003065913
training input 0.003033599
rate training 0.003027605
training methods 0.003026272
training example 0.003017975
translation 0.00300524
training frame 0.002954815
incremental training 0.00293926
training speed 0.0029363690000000003
training efficiency 0.002929644
ebuu training 0.0029255970000000003
cal training 0.002917214
training exam 0.002913448
cremental training 0.002909581
training effi 0.002907502
ious training 0.002907502
training 0.00264736
test data 0.0022316510000000003
evaluation data 0.001962846
development data 0.0018763550000000001
language model 0.0018388529999999999
data selection 0.001798797
data sets 0.001773995
retrieval data 0.001745399
word alignment 0.001731071
test sentence 0.001714404
local method 0.001611466
word sequence 0.001600518
local weight 0.0015739629999999998
test set 0.001524364
test sentences 0.001486602
following function 0.001477028
loss function 0.001473758
similar translations 0.001472226
lation model 0.001464293
level bleu 0.001430765
source sentences 0.001398846
algorithm mert 0.001369623
optimization algorithm 0.001357158
different sentences 0.001351725
bleu points 0.001351577
objective function 0.001312036
local weights 0.001303051
sentence pairs 0.001300963
single sentence 0.0012964869999999998
minus bleu 0.001293947
sentence level 0.001293458
similar source 0.001291176
global method 0.0012787620000000001
input sentence 0.001274242
oracle translations 0.001265866
target sentences 0.0012623410000000002
ing method 0.00125981
baseline weight 0.001259431
feature vector 0.001251299
local train 0.00121294
local methods 0.001205123
test example 0.001197016
