training data 0.0037612
training documents 0.002852875
training test 0.002791192
training set 0.002743705
test data 0.002669752
training examples 0.002484211
several data 0.002469007
labeled training 0.002430643
original training 0.002312086
ing data 0.002292854
training sets 0.002269138
training example 0.002239736
name training 0.002237788
data point 0.0022153010000000002
new data 0.002180285
data sets 0.002147698
complex data 0.0021368100000000003
separable data 0.002111567
free data 0.002101857
arbitrary data 0.0020983020000000002
data parts 0.0020983020000000002
extensive data 0.0020983020000000002
classification method 0.0020831499999999998
training 0.00194132
documents category 0.001848962
test documents 0.0017614269999999999
learning algorithm 0.0017301109999999999
category level 0.001670588
test set 0.0016522569999999999
text classification 0.001645578
category performance 0.0015473029999999999
other results 0.001521798
same category 0.0015103590000000002
hierarchical classification 0.0014890279999999999
classification problems 0.001482395
simple classification 0.001476984
second method 0.001463488
binary classification 0.001451136
machine learning 0.00144421
other methods 0.001417601
classification prob 0.001412015
practical classification 0.001409572
correct category 0.001407672
class level 0.0013915389999999998
online documents 0.001386039
classification systems 0.001382628
large feature 0.0013821039999999999
svms model 0.001378948
learning techniques 0.001369375
table category 0.001337612
distribution category 0.0013353240000000001
category distribution 0.0013353240000000001
chine learning 0.001327419
level hierarchy 0.00129305
negative examples 0.0012929109999999999
other models 0.001281001
ing set 0.001275359
level categories 0.001264313
category name 0.001233875
category hierarchies 0.001226785
other hand 0.001226757
category hier 0.001217889
other com 0.001211963
incorrect documents 0.001198667
documents nerror 0.001194517
nerror documents 0.001194517
unlabelled documents 0.001192017
second level 0.001183809
several methods 0.00115146
positive examples 0.001151091
optimization problem 0.00115083
feature spaces 0.001128935
binary classifier 0.001092682
bayes classifier 0.001085261
reuters corpus 0.001081864
classification 0.00107029
unique words 0.001068613
high performance 0.001067353
tistical problem 0.001065194
word independence 0.001056027
times word 0.0010552880000000001
learning 0.00104587
svm classifiers 0.0010372670000000001
top level 0.001028471
categorization accuracy 0.001021316
egory level 0.001017764
fourth level 0.001017577
third level 0.001014349
method 0.00101286
ing categories 0.001004106
probability value 0.0010034150000000001
accurate classifier 9.97702E-4
document assumption 9.91767E-4
model 9.88105E-4
vector space 9.87408E-4
large amount 9.87118E-4
aboost algorithm 9.65958E-4
examples figure 9.652700000000001E-4
document length 9.418709999999999E-4
category 9.37407E-4
