parsing model 0.002122788
first word 0.002074167
head word 0.001939679
dependency tree 0.001868425
single word 0.00178896
word bigrams 0.00177736
word form 0.001766098
bigram model 0.001748463
joint model 0.001708514
deletion model 0.001694259
tion model 0.001671643
subtree model 0.0016711669999999999
sion model 0.00164994
dependency parsing 0.001616239
parsing algorithm 0.001545916
dependency features 0.0014774789999999999
parse tree 0.001456916
time algorithm 0.001414697
model 0.00141011
first algorithm 0.001399985
algorithm first 0.001399985
dependency parse 0.001395613
sentence compression 0.001381351
dependency parser 0.0013803539999999999
language models 0.001367456
time parsing 0.001294137
dependency arc 0.001291745
optimal tree 0.001268887
programming algorithm 0.001246182
order dependency 0.00124478
dependency arcs 0.001232997
synchronous tree 0.001213478
tree substitution 0.001204713
second algorithm 0.0012021240000000002
tree struc 0.001201804
tree topolo 0.001196461
tree restriction 0.001196461
dependency type 0.001187193
virtual dependency 0.0011849389999999999
deleted words 0.0011818179999999998
ing algorithm 0.00118138
adjacent words 0.001176096
consecutive words 0.001175672
uninformative words 0.001170384
jective dependency 0.001148323
approximate algorithm 0.0011000110000000001
decent algorithm 0.0010959
optimization problem 0.001077172
compression corpus 0.001076614
recent parsing 0.0010680260000000001
algorithm enumer 0.001065779
original sentence 0.001063472
binary vector 0.001046623
feature vector 0.00102961
popular sentence 0.001009493
sentence compres 0.001009417
time complexity 0.0010090540000000001
dency parsing 0.001002164
compression evaluation 9.978769999999999E-4
labeling problem 9.93203E-4
other languages 9.90512E-4
sentence readable 9.903050000000001E-4
joint models 9.87672E-4
parsing techniques 9.71298E-4
compression task 9.62161E-4
space complexity 9.52466E-4
words 9.36743E-4
overall time 9.331389999999999E-4
compression rate 9.32016E-4
ing time 9.29601E-4
specific compression 9.2864E-4
objective function 9.233570000000001E-4
english penn 9.21591E-4
compression speed 9.124700000000001E-4
bigram features 9.12271E-4
dual variable 9.10952E-4
parse trees 9.083019999999999E-4
weight vector 9.04715E-4
dependency 9.03561E-4
final score 9.03093E-4
gold compression 9.01889E-4
learning task 8.909E-4
experimental results 8.87098E-4
new span 8.854640000000001E-4
compression ratio 8.725580000000001E-4
tence compression 8.72158E-4
other merges 8.720350000000001E-4
dency features 8.634039999999999E-4
treebank corpus 8.61927E-4
learning rate 8.60755E-4
first part 8.59491E-4
main assumption 8.57868E-4
unigram features 8.573879999999999E-4
space complex 8.51071E-4
same accuracy 8.49334E-4
exact decoding 8.36795E-4
feature weight 8.36619E-4
algorithm 8.33238E-4
total score 8.30716E-4
linear programming 8.29085E-4
