word sense 0.002751074
negative data 0.00248288
training data 0.0023909609999999996
positive sense 0.002385642
negative examples 0.0023475199999999996
sense prediction 0.002173167
sense instances 0.002164479
data set 0.002148424
correct sense 0.002147161
sense disambiguation 0.0021456590000000003
example data 0.002101039
sense wega 0.00209659
sense ambiguity 0.0020695460000000002
unlabeled data 0.0020592979999999998
negative training 0.0020496409999999996
learning method 0.002025688
active learning 0.002021452
sampling examples 0.001969254
unlabeled examples 0.001923938
learning approach 0.001919536
learning process 0.001909913
positive examples 0.0018793019999999998
learning technique 0.0018525689999999999
target data 0.001849052
percentage data 0.0018111199999999998
search data 0.0018007049999999999
data instances 0.0017934989999999998
sense 0.00178308
data instance 0.001762652
negative example 0.0017597189999999999
labeled examples 0.0017533289999999999
blog data 0.0017457879999999999
learning algo 0.0017254669999999998
training set 0.001715185
learning iterations 0.001714096
data sets 0.001713587
word senses 0.001634113
tive examples 0.001609853
wsd problem 0.0015923
negative dataset 0.001588238
different senses 0.001572534
evaluation word 0.001559766
wsd accuracy 0.0015239960000000001
training dataset 0.0014963189999999999
word feature 0.001465807
initial training 0.0014176199999999999
learning 0.00141093
target word 0.001404946
ambiguous word 0.001397178
classifier figure 0.001392849
word distribution 0.001392231
bayes classifier 0.001358618
binary classification 0.001346607
unlabeled example 0.0013361369999999998
sampling method 0.001307272
refined classifier 0.001302648
output classifier 0.001288975
examples 0.00127674
average wsd 0.001274506
building training 0.0012736219999999999
experimental results 0.001263817
classifier γem 0.001258642
backward word 0.001254117
random sampling 0.001250438
above results 0.001249862
ing method 0.00124256
initial accuracy 0.001238335
target words 0.001237087
ambiguous words 0.001229319
results figure 0.001223905
test set 0.001208199
related words 0.00118936
comparison results 0.0011786449999999999
problem setting 0.001176244
supervised wsd 0.001172197
human labeling 0.001171238
unlabeled dataset 0.001164656
initial wsd 0.001163179
high accuracy 0.00116292
same way 0.001158953
several example 0.001120688
positive dataset 0.00112002
upper accuracy 0.00111593
unsupervised wsd 0.001101662
wsd systems 0.0010994960000000002
ambiguous senses 0.001095303
wsd system 0.001090921
complete set 0.001067401
enough wsd 0.001066451
japanese web 0.001066139
other football 0.001051122
large number 0.001047338
simple text 0.001047311
uncertainty sampling 0.001039823
wsd accuracies 0.001035917
classification 0.00103574
text search 0.001024536
evaluation table 0.001024138
trainig set 0.001021671
web mining 0.001020853
