context model 0.001989838
dependency model 0.00188811
text model 0.001883813
unsupervised model 0.001875349
model accuracy 0.001845724
probabilistic model 0.001830315
regression model 0.0018030849999999998
span model 0.001791478
generative model 0.0017765379999999998
feature vector 0.001764524
model weights 0.0017588489999999998
model the 0.0017467189999999999
context features 0.0016358280000000002
model 0.00149534
feature name 0.0014400840000000001
span features 0.0014374680000000001
ccm training 0.001415064
basic feature 0.0014093320000000001
boundary feature 0.0013996430000000001
feature templates 0.001395785
intuitive features 0.001391639
binary tree 0.0013888289999999998
features fire 0.0013733270000000001
coarse features 0.0013733270000000001
large tree 0.001345108
dependency models 0.001162268
feature 0.00115878
tree fragments 0.001158707
unsupervised models 0.001149507
features 0.00114133
training corpus 0.0011283209999999998
standard ccm 0.001031511
ccm standard 0.001031511
ccm accuracy 0.0010268130000000001
ccm scores 0.001022851
constituent grammar 0.001012101
complex models 0.001000602
training sen 9.93264E-4
original ccm 9.88404E-4
baseline ccm 9.84349E-4
word segmentation 9.82911E-4
data sparsity 9.77886E-4
boolean matrix 9.76865E-4
constituent context 9.73653E-4
ccm right 9.702129999999999E-4
natural language 9.64346E-4
weight vector 9.61545E-4
binary trees 9.466170000000001E-4
probability mass 9.45771E-4
word alignment 9.4547E-4
the ccm 9.27808E-4
dependency grammar 9.25716E-4
positive probability 9.22024E-4
ccm llccm 9.19398E-4
learning procedure 9.181230000000001E-4
unsupervised grammar 9.12955E-4
ccm parame 9.08277E-4
ccm implementa 9.08277E-4
language processing 8.93829E-4
other distributions 8.7444E-4
joint distribution 8.688330000000001E-4
grammar induction 8.68428E-4
lexical information 8.615750000000001E-4
conditional distributions 8.525070000000001E-4
maximum sentence 8.469580000000001E-4
uniform distribution 8.42993E-4
pos tags 8.42662E-4
sentence length 8.409E-4
zero vector 8.401299999999999E-4
likelihood gradient 8.34432E-4
conditional probabilities 8.29626E-4
constituency parsing 8.24469E-4
recent research 8.10971E-4
constituent type 8.01648E-4
local probabilities 7.998109999999999E-4
several studies 7.97248E-4
complete sequence 7.89982E-4
objective gradient 7.83134E-4
constituent spans 7.82554E-4
constituent con 7.82144E-4
possible bracketings 7.75806E-4
dynamic program 7.73668E-4
models 7.69498E-4
cal distribution 7.64353E-4
context pair 7.630569999999999E-4
log likelihood 7.6158E-4
constituency parse 7.58819E-4
pos sequences 7.51578E-4
matrix 7.44512E-4
possible span 7.43899E-4
penn treebank 7.43403E-4
tive research 7.41219E-4
training 7.38635E-4
ing multinomials 7.378370000000001E-4
local distri 7.3741E-4
first unsuper 7.36581E-4
following example 7.33226E-4
machine translation 7.32567E-4
gold pos 7.28717E-4
context subsequences 7.25978E-4
