dependency parsing 0.00333561
parsing model 0.00292209
parsing algorithm 0.00290558
parsing state 0.0026118829999999997
parsing algorithms 0.002503955
same parsing 0.002419933
current parsing 0.002331363
parsing approach 0.002293789
parsing accuracy 0.002281919
projective parsing 0.002234172
dependency graph 0.002132768
dependency parsers 0.00212441
parsing experiments 0.002103765
greedy parsing 0.002103728
each parsing 0.00209921
initial parsing 0.002084366
parsing speed 0.002077704
parsing approaches 0.002072866
pendency parsing 0.002070231
jective parsing 0.002060012
ensemble parsing 0.002046745
dependency label 0.0019494809999999999
new transition 0.0019186429999999998
dependency trees 0.001884847
other transition 0.0018685949999999998
greedy dependency 0.001848538
current transition 0.0018370629999999999
dependency pars 0.00183349
feature vector 0.001809912
dependency representations 0.001798202
parsing 0.0017954
vanced dependency 0.001784615
different features 0.001772167
transition sequence 0.0017704219999999998
same model 0.001751223
same features 0.001656773
transition sequences 0.0016256019999999999
multiple transition 0.001621893
transition preconditions 0.0015966219999999998
buffer transition 0.0015925919999999999
bootstrapping transition 0.0015855329999999999
swap transition 0.0015504569999999999
dependency 0.00154021
previous model 0.001539795
ing algorithm 0.001533262
feature vectors 0.00150528
next model 0.0014894460000000002
feature map 0.001478044
feature extraction 0.001454773
feature vec 0.0014463
feature selection 0.001436929
feature engineering 0.0014361299999999999
subgradient algorithm 0.0014299540000000002
initial model 0.001415656
parse tree 0.001412836
tree approach 0.0014107780000000001
dimensional features 0.001385488
ensemble model 0.0013780350000000001
specific model 0.0013732000000000002
current state 0.001352446
different beam 0.001339024
transition 0.0013011
morphological features 0.001289924
unigram features 0.001287248
distance features 0.001281155
beam search 0.0012786770000000002
spanning tree 0.001250344
same beam 0.00122363
training data 0.0012122930000000001
greedy parser 0.001199335
feature 0.00118751
training set 0.001182519
search approach 0.001177969
final state 0.001176181
ing algorithms 0.001131637
model 0.00112669
same accuracy 0.001111052
algorithm 0.00111018
many transitions 0.0011065089999999999
initial state 0.001105449
linear time 0.001091128
state sij 0.001064827
other words 0.001060802
different num 0.00104811
linear programming 0.001047727
same number 0.001034008
features 0.00103224
training method 0.001014893
evaluation set 9.95197E-4
different margins 9.90914E-4
different kinds 9.89266E-4
same values 9.89035E-4
other label 9.76766E-4
projective dependencies 9.495630000000001E-4
new sequences 9.420450000000001E-4
same margin 9.35708E-4
pos tag 9.354960000000001E-4
beam size 9.24778E-4
grid search 9.246360000000001E-4
development set 9.21204E-4
