argument word 0.00330176
syntactic features 0.002458353
word path 0.0023901549999999997
srl model 0.00236425
word sequence 0.002265688
word span 0.0022622099999999997
hmm features 0.002239622
word spans 0.002211487
word sense 0.002203648
single word 0.00219843
word paths 0.002175008
explicit word 0.0021424029999999998
average word 0.002121865
preceding word 0.002104717
path features 0.002059395
first argument 0.002022148
hmm model 0.001973762
argument identification 0.001971189
baseline features 0.0019664929999999997
following features 0.001930338
argument pos 0.00192846
argument classification 0.001904096
argument phrase 0.001858888
useful features 0.001844704
language model 0.001833421
following argument 0.001831118
ing argument 0.001830559
argument type 0.001822699
potential argument 0.001799924
categorical features 0.001786686
last argument 0.001781097
sophisticated features 0.001777954
srl results 0.001765945
test data 0.001717147
baseline model 0.001700633
srl system 0.001689281
argument msh 0.001685915
argument identifi 0.0016790379999999999
argument classi 0.001674981
ing model 0.001663919
training data 0.0016556869999999999
domain srl 0.001643543
crf model 0.001598429
srl task 0.001570471
sparse model 0.001568179
bayes model 0.001565575
data set 0.001547363
standard srl 0.001546895
features 0.00153511
baseline srl 0.001526383
cal model 0.0015164459999999999
sophisticated model 0.001512094
supervised srl 0.001511344
basebylength model 0.00150945
abilistic model 0.001507852
corpus data 0.001493189
semantic role 0.001490177
srl systems 0.0014782200000000001
new feature 0.001474869
unsupervised srl 0.001466616
srl pipeline 0.0014560620000000002
different corpus 0.001437287
hmm models 0.001436672
argument 0.00143589
srl experiment 0.001428549
test set 0.001409724
feature space 0.001401967
feature values 0.001387989
top srl 0.001364556
srl sys 0.001357811
corpus test 0.00135555
test corpus 0.00135555
training set 0.001348264
different classification 0.001339697
portable srl 0.001333233
models system 0.001326441
suitable feature 0.001325569
ing data 0.001322062
test predicates 0.0013194460000000002
hmm system 0.001298793
unlabeled data 0.0012970500000000001
language models 0.001296331
labeled data 0.001273941
model 0.00126925
hmm language 0.001268683
separate data 0.001262847
hmm state 0.001260698
state hmm 0.001260698
data points 0.00125515
data point 0.001247674
hmm path 0.001228797
sparse data 0.001226322
different spans 0.001217108
data sparsity 0.001208081
predicate identification 0.001206951
training examples 0.001206894
same hmm 0.001201949
base hmm 0.001184546
many words 0.00118418
syntactic relationships 0.001182757
