training words 0.00225527
training data 0.002055903
training examples 0.0019252870000000001
probability model 0.00186132
training corpus 0.001819303
training set 0.0017962250000000003
training methods 0.0017591020000000002
model size 0.001741457
other words 0.001720905
single model 0.0017071999999999999
probabilistic model 0.0017037369999999999
model parameters 0.001701348
markov model 0.00168531
different selection 0.001641012
hmm model 0.001638654
training counts 0.001607085
bigram model 0.001594367
particular model 0.001592582
current training 0.00159004
trained model 0.001586023
training cost 0.0015831530000000001
supervised training 0.001577513
model parame 0.001553605
unsupervised training 0.001541209
training corpora 0.001495901
several words 0.001478343
complete training 0.001475553
selection method 0.001474689
training natu 0.0014687810000000002
selective training 0.0014628660000000002
word sense 0.001459227
low training 0.0014587240000000002
word sequence 0.001458621
plete training 0.001456218
head word 0.0014537810000000001
labeling word 0.001449109
ion words 0.001426951
ambiguous word 0.001409338
different methods 0.001407311
learning method 0.001374956
model 0.00133293
language models 0.001296372
selection probability 0.001282013
selection methods 0.001273545
ambiguous words 0.001267918
possible tag 0.00125499
unambiguous words 0.0012476120000000002
training 0.00123918
biguous words 0.001233893
example selection 0.001216811
selection approach 0.001215401
selection work 0.001201362
sample selection 0.001182403
different committee 0.001180944
selection algorithm 0.001175557
different batch 0.001162
other work 0.001152554
tag probabilities 0.001140903
new examples 0.001133922
different type 0.001127734
conditional probability 0.001126075
different numbers 0.001124886
probability distribution 0.0011102389999999998
different variants 0.0011094240000000001
different multinomials 0.001109345
sampling method 0.001099095
tag sequence 0.001094918
tion method 0.0010929820000000002
learning algorithm 0.001075824
many examples 0.0010715310000000001
tag sequences 0.0010670480000000001
test set 0.001059934
such counts 0.001052368
selection system 0.001051055
sequential selection 0.001049425
similar examples 0.00104884
tagging accuracy 0.001047812
natural language 0.001046664
such modeling 0.001045322
example distribution 0.001045037
such tasks 0.001040008
same accuracy 0.001038521
labeled examples 0.001037655
language processing 0.001037545
selection criteria 0.001034861
data sparseness 0.001033935
selection algorithms 0.0010304300000000001
batch selection 0.001028234
selection scheme 0.00102629
eral selection 0.001023811
words 0.00101609
unlabeled examples 0.001014358
alternative selection 0.001013081
tag occurring 0.001011548
other score 0.001009249
linear selection 0.001001728
several parameter 9.98384E-4
statistical methods 9.96474E-4
member selection 9.903800000000001E-4
classification accuracy 9.85044E-4
