training data 0.0028954899999999997
test data 0.002467146
data set 0.002465465
unlabeled data 0.002441078
large data 0.0023364099999999997
labeled data 0.002293266
feature set 0.002272025
ing data 0.00226094
data sets 0.002242063
chunking data 0.002218636
ner data 0.002198321
different feature 0.002182036
development data 0.002136836
model parameters 0.002129962
feature vector 0.0020921
data points 0.002074391
several feature 0.002068243
incomplete data 0.002065705
nearby data 0.002065705
whole data 0.002065705
additional feature 0.002053597
generative model 0.002048712
feature sets 0.002048623
model structures 0.002038257
feature types 0.002019094
markov model 0.001992791
appropriate model 0.001986583
hybrid model 0.001975701
ture model 0.001959668
feature type 0.001958817
model parame 0.001941962
mixture model 0.001935298
brid model 0.001934006
feature engineering 0.0019252219999999999
same training 0.001901481
feature design 0.001880428
feature division 0.001873383
model 0.00168881
crf training 0.0016773539999999998
output learning 0.001659405
feature 0.00162759
discriminative training 0.0015660639999999998
labeled training 0.001546696
baseline features 0.0015350899999999998
probability models 0.001509204
objective function 0.001487552
supervised learning 0.001457677
learning tasks 0.001453203
sequence learning 0.001435052
discriminative models 0.001428179
partition function 0.0014157380000000002
training hysol 0.001404594
training criterion 0.001403599
tive function 0.001402443
discriminant function 0.00139947
training procedure 0.001393595
potential function 0.001392806
same way 0.001390369
same optimization 0.00138089
cave function 0.001348837
generative models 0.001296477
learning setting 0.001293408
word type 0.001280492
learning settings 0.001275811
word table 0.001265918
probabilistic models 0.001253934
individual models 0.001236386
tive models 0.001235908
classification performance 0.001230006
same chunking 0.001224627
sentence accuracy 0.001219198
quence learning 0.001218778
graphical models 0.0012168090000000001
optimization algorithm 0.001213738
same objective 0.001211463
word prefix 0.001209651
good performance 0.001190052
ative models 0.0011900510000000001
criminative models 0.001185094
powerful models 0.001181661
prior probability 0.001152671
supervised set 0.00112981
estimation algorithm 0.001128219
conditional probability 0.00112441
prior distribution 0.001116699
same manner 0.001113702
probability distribution 0.001109286
function 0.00110311
features 0.00110152
ner performance 0.001092384
output space 0.001089388
same class 0.0010863140000000001
training 0.00107446
output label 0.00107412
same man 0.001073414
output values 0.001067063
discriminative approach 0.001052523
hysol performance 0.001045227
ing log 0.001035514
algorithm let 0.001029947
