language word 0.00316959
language tags 0.0025883900000000003
current model 0.002523562
new model 0.002481831
previous model 0.0024506529999999997
original model 0.002411493
seed model 0.002392313
model coverage 0.00237285
ture model 0.002351271
ﬁnal model 0.0023378319999999998
ibm model 0.0023326989999999997
target language 0.002330507
gram model 0.002320237
source language 0.002305074
tag information 0.002240985
training data 0.00219746
tag set 0.0021838279999999996
tag dictionary 0.002176321
model 0.00206304
tag distribution 0.0020094659999999997
tag projection 0.00200047
unsupervised tag 0.00199252
related language 0.001971695
word align 0.001923491
english pos 0.001840927
word wtj 0.0018290390000000002
word wsi 0.0018290390000000002
tag infor 0.0017802089999999998
target languages 0.0017774470000000001
uniform tag 0.001777018
parallel data 0.0017738520000000002
training sentences 0.0017687039999999998
other languages 0.001711256
training corpus 0.0016513919999999998
test data 0.001642028
pos tagger 0.0016281450000000001
data tokens 0.001613934
language 0.00159847
same tags 0.001568713
french data 0.0015057850000000002
supervised pos 0.0014752020000000001
little data 0.00145767
labelled data 0.001432396
unsupervised pos 0.001392926
random pos 0.001369798
pos taggers 0.0013606900000000001
target corpus 0.0013575290000000001
transfer information 0.001337758
training iterations 0.001335058
sentence alignment 0.001329447
universal tags 0.001326792
ual languages 0.001321874
self training 0.001321269
parallel source 0.001308896
lar languages 0.001308434
romance languages 0.0013041580000000001
germanic languages 0.001303101
pos tagging 0.001302344
project tags 0.001298582
labelled training 0.001286736
basic tags 0.001261395
conﬁdence tags 0.001247654
target lan 0.001242642
source algorithm 0.001241677
parallel corpus 0.001227784
case alignments 0.001227677
source lan 0.001217209
other methods 0.001215278
preliminary english 0.0011925899999999999
unsupervised tagger 0.0011855989999999999
english laws 0.001180763
melt pos 0.00117894
stanford pos 0.001175515
unaligned words 0.001134759
new tagger 0.0011292
hmm tagger 0.001108873
high accuracy 0.0011009449999999999
previous tagger 0.001098022
unknown words 0.001092794
guage sentences 0.00109168
average accuracy 0.001089063
known words 0.001081868
overall accuracy 0.001079928
transition probabilities 0.00107004
foreign words 0.0010684129999999998
ing corpus 0.001068405
ﬁdence alignments 0.00106739
transition probability 0.0010589739999999999
alignment score 0.001058657
target side 0.0010578990000000002
languages 0.00104541
french tagger 0.001044634
unsupervised approach 0.001043933
syntactic similarity 0.001043361
seed tagger 0.0010396819999999999
parallel texts 0.001032677
large number 0.001030192
our tagger 0.00103002
feature 0.00102689
low accuracy 0.001026384
