decoding algorithm 0.001562296
training algorithm 0.00154222
same features 0.001537093
viterbi algorithm 0.001475822
same time 0.001450313
search algorithm 0.001436525
ing algorithm 0.001379254
decoding time 0.00133584
label features 0.001325237
syntactic parsing 0.0013230590000000001
training time 0.001315764
exact algorithm 0.001298987
efficient algorithm 0.0012899839999999999
different decoding 0.001251082
carpediem algorithm 0.001226243
common features 0.001216165
algorithm calls 0.0012113739999999999
search time 0.001210069
perceptron algorithm 0.00119161
quence algorithm 0.00118637
tron algorithm 0.001180441
time complexity 0.00117759
different tag 0.001175506
training set 0.00117504
such tag 0.0011590160000000001
standard features 0.001153652
training data 0.00115325
tag decoding 0.001139478
viterbi decoding 0.00112858
based features 0.001114041
specific features 0.001104299
search decoding 0.001089283
feature size 0.0010858110000000001
parsing context 0.001083933
input sequence 0.001080389
sequence score 0.001078605
decoding algorithms 0.001074434
decoding problem 0.001064501
time com 0.001058775
unigram features 0.001057054
training algorithms 0.001054358
search data 0.001047555
bigram features 0.001042126
computation time 0.0010405829999999999
same way 0.001036855
real time 0.0010314299999999998
same accuracy 0.001031007
pcfg parsing 0.001023971
pos tag 0.001019949
edge score 0.0010168970000000001
feature functions 0.00100835
semantic labels 9.92526E-4
response time 9.92301E-4
sequence tagging 9.88957E-4
arbitrary feature 9.88369E-4
different heuristics 9.84577E-4
first viterbi 9.683509999999999E-4
current word 9.67735E-4
feature patterns 9.642730000000001E-4
feature engi 9.642730000000001E-4
same manner 9.62332E-4
node score 9.60468E-4
additional time 9.568199999999999E-4
algorithm 9.54769E-4
different degenerate 9.53732E-4
sequence size 9.49625E-4
decoding context 9.354859999999999E-4
beam decoding 9.33292E-4
tion function 9.253989999999999E-4
decoding algo 9.24055E-4
other ones 9.222550000000001E-4
sequential decoding 9.218729999999999E-4
active sequence 9.15328E-4
viterbi case 9.09263E-4
large label 8.99937E-4
decoding problems 8.997479999999999E-4
decoding speed 8.969029999999999E-4
such pruning 8.91942E-4
current node 8.91849E-4
test set 8.91712E-4
ing algorithms 8.91392E-4
corresponding sequence 8.826999999999999E-4
new nodes 8.8147E-4
ing problem 8.81459E-4
pos tagging 8.76808E-4
iterative viterbi 8.74755E-4
different positions 8.73126E-4
optimal sequence 8.725989999999999E-4
perceptron model 8.69508E-4
approximate decoding 8.67401E-4
decoding process 8.667289999999999E-4
sequence position 8.62678E-4
return sequence 8.61776E-4
crfs model 8.615109999999999E-4
label size 8.59622E-4
tagging algorithms 8.55717E-4
crf training 8.512400000000001E-4
pos tags 8.508350000000001E-4
total decoding 8.50627E-4
ceiling function 8.47461E-4
