corpus method 0.00165301
same data 0.001556965
other corpus 0.001556456
first corpus 0.00152433
ing data 0.001467681
parse tree 0.0014671950000000001
method language 0.001406127
structure parse 0.00140022
extraction model 0.0013960029999999998
lexical features 0.001370478
relation corpus 0.001366788
test data 0.001348719
token model 0.001335847
ing method 0.001323369
parse nodes 0.001303468
second corpus 0.001279476
constituent model 0.001270507
model comparison 0.001270324
relations corpus 0.001269698
parc corpus 0.001249694
accuracy results 0.001240349
doc corpus 0.001239015
corpus doc 0.001239015
following features 0.001238942
stituent model 0.001235676
herald corpus 0.00123331
crf model 0.00122888
smhc corpus 0.001227707
language test 0.0012221060000000001
corpus sen 0.001207802
span features 0.0012006080000000001
common features 0.001199233
test set 0.0011989700000000002
information extraction 0.00119023
gold parse 0.001185119
sequence features 0.001172434
parse french 0.001146738
full results 0.001137589
unlabelled data 0.0011124210000000002
development data 0.00110155
pendency parse 0.001096402
sequence information 0.001092256
scale data 0.001082569
full set 0.00106729
overall results 0.001061978
size results 0.001059878
original word 0.001058815
final results 0.001055959
attribution methods 0.0010501669999999999
token method 0.001037777
dependency 0.00103541
other models 0.001031449
smhc results 0.001027987
token approach 0.001026024
second method 0.001014894
structure nodes 0.0010066720000000001
typewise results 9.960120000000001E-4
parable results 9.960120000000001E-4
model 9.92284E-4
final method 9.91097E-4
different characters 9.89592E-4
large news 9.89121E-4
tion methods 9.86627E-4
direct quotation 9.77605E-4
different languages 9.77145E-4
constituent method 9.72437E-4
large corpora 9.71894E-4
constituent approach 9.606840000000001E-4
corpus 9.58796E-4
different degrees 9.56557E-4
pos tags 9.48499E-4
such speech 9.46937E-4
future methods 9.44037E-4
factored parser 9.427299999999999E-4
quotation attribution 9.38818E-4
third method 9.3611E-4
effective method 9.32612E-4
quotation extraction 9.32352E-4
extraction quotation 9.32352E-4
method dir 9.312449999999999E-4
ing process 9.296069999999999E-4
closed set 9.29007E-4
phrase structure 9.27353E-4
direct quotations 9.16498E-4
text portuguese 9.1336E-4
ing herald 9.03669E-4
large class 9.0334E-4
attribution accuracy 8.914579999999999E-4
text mining 8.90161E-4
other cor 8.85628E-4
simple system 8.853999999999999E-4
previous work 8.80487E-4
quotations table 8.78911E-4
syntactic patterns 8.75238E-4
training 8.71151E-4
quotation type 8.70938E-4
attribution system 8.69504E-4
features 8.66689E-4
quotation marks 8.664460000000001E-4
quotation span 8.62552E-4
