whole word 0.001717997
word occurrences 0.001688941
training data 0.0016000300000000001
tree model 0.0015027830000000002
model parameter 0.001467328
model distribution 0.001440852
ing data 0.0014321730000000001
model parameters 0.0013698500000000001
other words 0.001362658
data selection 0.001313093
data figure 0.00131153
training example 0.001295137
different words 0.001284977
hmm model 0.0012818270000000001
training corpus 0.001256667
training set 0.001248531
probability distribution 0.001243116
markov model 0.001240634
tree algorithm 0.0012302720000000001
data sparseness 0.001103274
data amount 0.00109824
sampling method 0.001097874
active data 0.001084261
training classifiers 0.001067851
input words 0.001054751
tree learning 0.001054288
new approach 0.001023912
frequency words 0.001020731
cue feature 0.001008132
different results 0.001007508
transformation method 0.001004109
particular method 9.95959E-4
sampling approach 9.94841E-4
rior probability 9.92967E-4
other approaches 9.778640000000002E-4
current sentence 9.75755E-4
model 9.74949E-4
partit ion 9.73027E-4
training sam 9.46508E-4
training blocks 9.45155E-4
cbs method 9.444599999999999E-4
final sentence 9.34429E-4
present method 9.3392E-4
method fares 9.3392E-4
novel method 9.3392E-4
first class 9.2206E-4
decision tree 9.16784E-4
other disagreement 9.164170000000001E-4
such models 9.14435E-4
discourse information 9.14275E-4
initial sentence 9.021109999999999E-4
selection function 8.99256E-4
parameter values 8.926720000000001E-4
input example 8.89686E-4
last sentence 8.8103E-4
tree models 8.67111E-4
learning models 8.65731E-4
tree classifiers 8.652200000000001E-4
total number 8.61096E-4
other hand 8.600630000000001E-4
learning context 8.582609999999999E-4
sentence ndings 8.52887E-4
sampling distribution 8.460399999999999E-4
cbs approach 8.41427E-4
preceding sentence 8.39908E-4
large disagreement 8.37228E-4
possible values 8.34439E-4
hmm tag 8.20814E-4
available information 8.14993E-4
records information 8.06576E-4
useful example 8.06352E-4
particular learning 8.04676E-4
number indi 8.03992E-4
multiple learning 8.00438E-4
preferred example 7.80966E-4
local distribution 7.794060000000001E-4
large vari 7.782700000000001E-4
tree classifier 7.77606E-4
probability 7.77213E-4
sufficient number 7.75133E-4
ordinal number 7.74903E-4
tree strategy 7.72743E-4
same time 7.72383E-4
large variance 7.72376E-4
large volume 7.715830000000001E-4
large propor 7.715830000000001E-4
first column 7.70756E-4
number indicat 7.691180000000001E-4
machine learning 7.67511E-4
first author 7.61746E-4
relation type 7.576390000000001E-4
probabilistic classifiers 7.548E-4
learning algo 7.51388E-4
class relation 7.49276E-4
normal distribution 7.458569999999999E-4
learning strat 7.42711E-4
tree classi 7.42596E-4
discrete value 7.342500000000001E-4
info value 7.33287E-4
training 7.30465E-4
