parsing model 0.0019592639999999996
parsing time 0.001620706
statistical parser 0.001581643
pcfg parser 0.001504919
stochastic parser 0.001490879
local information 0.0014709509999999999
simple model 0.001449464
different gap 0.00143975
perfect model 0.001410216
ing model 0.001398662
linear information 0.001359345
structure information 0.001354056
perfect information 0.001349941
syntactic information 0.001345507
main results 0.001313828
unlexicalized model 0.0013092029999999999
different sources 0.0013078389999999999
pcfg model 0.0012985989999999999
different ways 0.0012932479999999999
probabilistic model 0.0012917509999999998
different condi 0.001289734
different module 0.001288568
combined model 0.001284724
parsing experiments 0.001280736
average parsing 0.001280554
insert model 0.001274126
noinsert model 0.001270023
parsing algorithms 0.001256447
prior information 0.001248415
parser 0.00122332
complicated parsing 0.0011902619999999999
training corpus 0.0011892769999999999
conditional parsing 0.001180834
efficient parsing 0.001177904
fore parsing 0.0011768389999999998
word level 0.001158759
recovery results 0.001133853
new method 0.001131095
such dependencies 0.001110535
development data 0.001096386
test set 0.001042927
such sites 0.001039918
training algorithm 0.001035434
sentence sam 0.0010279669999999999
word sam 0.001021763
recovery accuracy 0.001020878
model 0.001017
local features 0.0010087479999999998
adjacent words 0.00100124
rate words 9.98893E-4
same problem 9.97817E-4
ing corpus 9.87305E-4
combination method 9.632339999999999E-4
tagging accuracy 9.592649999999999E-4
perfect case 9.58637E-4
information 9.56725E-4
current sentence 9.55045E-4
arate words 9.49302E-4
new label 9.4421E-4
parsing 9.42264E-4
treebank grammar 9.35209E-4
word sasha 9.275189999999999E-4
state method 9.254649999999999E-4
low tagger 9.22834E-4
time increases 9.17699E-4
such constructions 9.04428E-4
other hand 9.03258E-4
previous work 8.91808E-4
such phenomena 8.868929999999999E-4
evaluation purposes 8.84513E-4
evaluation mea 8.84513E-4
other terms 8.832639999999999E-4
probability models 8.77514E-4
standard test 8.76958E-4
main contributions 8.73069E-4
shallow models 8.63405E-4
iterative training 8.613329999999999E-4
gap feature 8.59258E-4
test section 8.58915E-4
antecedent recovery 8.58382E-4
recovery algorithm 8.53124E-4
wsj corpus 8.49179E-4
local cues 8.40297E-4
main difference 8.33126E-4
semantic analysis 8.30973E-4
recent work 8.28919E-4
insert case 8.22547E-4
main bottleneck 8.18615E-4
main aim 8.18615E-4
beam search 8.151079999999999E-4
possible sites 8.13777E-4
trace tagger 8.12746E-4
binary feature 8.11904E-4
related work 8.0408E-4
annotation errors 7.95683E-4
low probability 7.93073E-4
important limitation 7.91895E-4
important concern 7.91072E-4
syntactic structure 7.86113E-4
first question 7.84992E-4
