data annotation 0.0029374049999999997
training data 0.002857832
data set 0.002817316
labeled data 0.002675351
ing data 0.0026661819999999996
data quality 0.002643446
data noise 0.0026098239999999997
unlabeled data 0.002584956
data points 0.0024517439999999996
data subset 0.0024427
ter data 0.002413508
data point 0.002397733
twitter data 0.0023927949999999996
testing data 0.002389634
data anno 0.0023710629999999996
noisy data 0.002361425
data cleaning 0.002352851
different class 0.0018469720000000001
learning algorithm 0.001833333
class label 0.001731043
annotation label 0.001626286
classification performance 0.0016089659999999999
words feature 0.001562302
training label 0.001546713
class examples 0.0015461609999999999
svm classification 0.001534958
training instances 0.0015165909999999999
feature vector 0.001469364
training set 0.001458488
classification methods 0.001448754
same class 0.0014420499999999998
such features 0.001441178
feature scores 0.00143494
learning process 0.001429204
ing algorithm 0.00142626
active learning 0.0014182700000000001
different number 0.001416399
machine learning 0.0013844320000000001
many instances 0.001375187
learning iterations 0.0013677890000000001
words svm 0.001366781
annotation quality 0.0013641909999999998
different classes 0.00136279
learning datasets 0.00135782
learning algorithms 0.0013533640000000001
labeled instances 0.00133411
quality label 0.001332327
classification model 0.001329028
learning approach 0.001327123
class distribution 0.001323546
labeled training 0.001316523
learning framework 0.001314548
emotion annotation 0.001299936
feature weights 0.001299854
label noise 0.001298705
annotation process 0.001293354
svm classifiers 0.0012918629999999999
feature bias 0.001283874
minority class 0.001283092
learning models 0.001281583
learning procedure 0.001279989
different approach 0.001275338
labeled dataset 0.001272094
feature weighting 0.0012676229999999998
classification task 0.001263344
active label 0.001250556
different subset 0.00124751
other words 0.0012471000000000001
feature level 0.001247036
majority class 0.001244149
quality dataset 0.001240189
dataset quality 0.001240189
scheme different 0.001231348
emotion classification 0.001230609
tive learning 0.0012297240000000002
training instance 0.001228717
selection methods 0.001220636
cheap feature 0.0012182019999999998
feature vectors 0.001216705
classification accuracy 0.001216275
same dataset 0.001213291
different strategies 0.001200842
new labels 0.00119634
performance comparison 0.001194201
specific features 0.001192472
spreading algorithm 0.001191114
average training 0.001185046
imbalanced class 0.001182247
language annotation 0.001180337
emotion dataset 0.001175934
ensemble classifier 0.001173887
discriminative features 0.001173632
different emotions 0.001173319
quality labels 0.001167757
annotation errors 0.001155967
spread classifier 0.001154281
algorithm delta 0.001154121
method spread 0.001153538
effective features 0.001152733
classification algorithms 0.001148187
