parse tree 0.002247958
tree set 0.002012425
same tree 0.002006149
binary tree 0.0018906449999999998
parsing model 0.0018859649999999999
subtree probability 0.001850449
random tree 0.0018271029999999998
tree kernel 0.0018091829999999998
dop model 0.001768047
binarized tree 0.0016763359999999998
parse trees 0.001668044
unsupervised model 0.001574845
parsing algorithm 0.001545215
other subtrees 0.001539781
same parse 0.001508047
supervised model 0.001497893
learning model 0.00144886
set trees 0.001432511
model english 0.001405973
combined model 0.0014044069999999999
corpus subtrees 0.001395446
same node 0.0013949090000000002
probability distribution 0.001387105
same data 0.001381763
ccm model 0.001381034
training data 0.001379143
model limit 0.001378242
first subtree 0.001377821
total probability 0.0013489650000000001
wsj trees 0.0013349529999999998
semantic roles 0.001325186
binary trees 0.0013107309999999999
different sentence 0.001296036
different parsing 0.0012736280000000002
possible trees 0.001271386
training set 0.001269894
dop parsers 0.0012682779999999999
forward probability 0.001263735
this subtree 0.001244236
other derivations 0.00124278
second subtree 0.001242435
subtree probabilities 0.001236866
probable parse 0.001225233
parse accuracy 0.001223528
unsupervised parsing 0.00121871
new sentence 0.001218132
parsing models 0.001194881
syntactic structure 0.001173079
parsing technique 0.00116192
language models 0.001142222
supervised parsing 0.001141758
empirical results 0.0011412409999999999
following subtrees 0.00113574
different parsers 0.001129994
covering subtree 0.001124688
model 0.00112105
test set 0.001112302
annotated data 0.001111788
unlabeled data 0.0011094120000000002
dop approach 0.0011087789999999998
large number 0.001101978
unsupervised dop 0.001100792
entire trees 0.00109731
structural language 0.001089634
parsing methods 0.0010893
root node 0.001084199
frontier node 0.0010837660000000001
susan other 0.001078897
dop models 0.001076963
unsupervised parsers 0.001075076
treebank pcfg 0.0010744349999999999
sentence investors 0.001063438
other half 0.001057138
frequent trees 0.001056699
internal node 0.001055679
unsupervised parser 0.001049974
wsj sentences 0.001041606
disappointing results 0.001040161
hypergraph parsing 0.001036313
sample subtrees 0.001035452
syntactic categories 0.001019693
same method 0.001017483
annotated corpus 0.001015785
raw data 0.0010157550000000001
unannotated data 0.001005641
total set 0.001005162
total number 0.001001829
reestimated subtrees 0.0010013980000000001
same evaluation 9.99782E-4
supervised parsers 9.98124E-4
probability 9.83198E-4
dependency models 9.73312E-4
supervised parser 9.73022E-4
first time 9.729910000000001E-4
unsupervised version 9.71628E-4
different context 9.67882E-4
top dop 9.61758E-4
same root 9.55528E-4
unlimited training 9.411619999999999E-4
previous work 9.403980000000001E-4
