sentence compression 0.00464481
human compression 0.003014708
source sentence 0.002942026
first sentence 0.002929498
compression system 0.002915776
original sentence 0.002907329
sentence com 0.002898441
other compression 0.0028887550000000002
sentence output 0.002848767
output sentence 0.002848767
sentence compres 0.002811736
english sentence 0.002809351
compression performance 0.002801927
input sentence 0.002798952
compression method 0.002783165
japanese sentence 0.002760171
reference sentence 0.002750331
lead sentence 0.002744617
conventional sentence 0.002739198
sentence 推定し 0.002728025
sentence compressor 0.002717315
sentence class 0.002716766
compressed sentence 0.00271171
oriented sentence 0.002711626
demand sentence 0.00271001
sentence align 0.002707926
compression rate 0.002543854
compression methods 0.002498104
sentence 0.00248315
compression compound 0.0024395370000000003
japanese compression 0.0024386810000000003
reference compression 0.002428841
shiken compression 0.0023998400000000003
compression rates 0.0023996160000000002
compression meth 0.002392304
erence compression 0.002389777
proved compression 0.002388479
moderate compression 0.002386567
compression 0.00216166
word dependency 0.001826233
other words 0.0016632119999999999
language model 0.0016584350000000002
syntactic information 0.0016423000000000002
tive model 0.001485942
original sentences 0.001457109
word pairs 0.00144984
inative model 0.001443814
dependency tree 0.00143902
ear model 0.0014162900000000002
generative model 0.0014142710000000001
guage model 0.0014130260000000002
test sentences 0.001404714
sentences output 0.001398547
word sequence 0.0013870520000000002
content words 0.001387034
news corpus 0.001386813
word bigram 0.0013627160000000001
english sentences 0.001359131
input sentences 0.001348732
grammatical sentences 0.001345597
text corpus 0.001330289
new features 0.001324983
word dependencies 0.0013126280000000001
japanese sentences 0.0013099510000000002
long sentences 0.0012977560000000002
lead sentences 0.001294397
syntactic tree 0.001292357
structural information 0.001286505
general corpus 0.001259849
dependency probability 0.001246865
test corpus 0.00123294
significance score 0.001218306
dependency relation 0.001218129
text data 0.001207121
arbitrary words 0.0011995909999999999
adjacent words 0.001199524
drop words 0.001197951
model 0.00118306
syntactic constraints 0.001182433
other content 0.001178012
bleu score 0.001173102
dependency trees 0.001168405
mean score 0.001164254
cent words 0.0011623599999999999
functional words 0.0011623599999999999
dependency parser 0.001155208
corpus size 0.0011364819999999999
novel features 0.001133361
huge corpus 0.001128294
summarization systems 0.001126766
human evaluations 0.001119756
dependency relations 0.0011171
human subjects 0.0011159310000000001
combined features 0.001112116
test data 0.001109772
original tree 0.0010839460000000001
evaluation label 0.00108216
japanese dependency 0.001056274
tree structure 0.001047017
feature set 0.00104578
