training algorithm 0.00259708
feature set 0.002550942
training data 0.00244123
tree algorithm 0.00242387
feature vector 0.002401606
training set 0.002398752
parse tree 0.002385973
training sentence 0.002257571
word features 0.002256822
similar feature 0.002233454
feature space 0.002141706
training sentences 0.002133157
other features 0.002127109
full feature 0.002116502
training time 0.0021161360000000002
feature vectors 0.002102417
training examples 0.002088286
dimensional feature 0.002048176
same tree 0.001967852
tree kernel 0.001966616
eral training 0.001937237
whole training 0.001929729
tence training 0.001904777
training points 0.001896688
training phase 0.001884448
correct tree 0.0018399290000000001
explicit features 0.001814505
perceptron algorithm 0.001814016
feature 0.00178425
journal tree 0.001781859
tree fragments 0.001719357
likely tree 0.001718472
tree fragment 0.001712978
able tree 0.001712669
training 0.00163206
probability model 0.001581081
parse trees 0.0015701259999999998
baseline model 0.001557601
features 0.00155646
parameter vector 0.001524878
different parameter 0.001496382
test data 0.001453359
example parse 0.001438576
simple algorithm 0.0014117790000000002
test set 0.0014108810000000001
algorithm figure 0.00139602
sentence set 0.001392203
form algorithm 0.00137846
parameter estimation 0.0013751660000000001
original model 0.001364546
dop model 0.001352801
probabilistic model 0.001338978
original algorithm 0.001321096
same data 0.001318172
programming algorithm 0.001313789
initial model 0.001298825
test sentence 0.0012697
dual algorithm 0.0012578720000000002
equivalent algorithm 0.001254268
perceptron algorithms 0.001247831
parameter values 0.001220186
same word 0.001209364
original perceptron 0.001205072
rect parse 0.001184232
own parse 0.0011812939999999998
parameter vec 0.0011682300000000001
parameter weighting 0.00116251
parameter esti 0.001160449
test example 0.0011556420000000001
set size 0.001143332
perceptron algo 0.001134668
perceptron meth 0.001116991
voted perceptron 0.0011081799999999998
new approach 0.0011048339999999999
web data 0.001103325
test examples 0.001100415
data sets 0.001091432
sparse data 0.001090296
previous word 0.001086145
development data 0.001082624
parsing methods 0.001069285
ble data 0.001067393
same number 0.001065404
representative set 0.001058869
different method 0.001058641
learning task 0.001056866
exact parsing 0.001054119
development set 0.001040146
relative error 0.001035594
finite set 0.001028656
experimental results 0.00102511
ment set 0.001020757
new learning 0.001015282
many algorithms 0.001014924
final test 0.001014278
model 0.00100847
vector representations 0.0010026969999999999
error rate 9.96584E-4
different algorithms 9.87695E-4
different methods 9.86248E-4
