learning algorithm 0.0019258079999999998
search algorithm 0.001808806
feature set 0.001751867
same feature 0.001696118
data search 0.001676565
ing algorithm 0.001557193
data set 0.001556696
training data 0.001514371
test data 0.001509307
same data 0.0015009469999999999
feature vector 0.0014788470000000001
context features 0.001452528
learning problem 0.001396423
left word 0.001394806
sequence learning 0.0013788519999999999
few feature 0.0013704
standard data 0.001361857
different features 0.001345394
new learning 0.001345155
feature sets 0.0013354550000000001
feature templates 0.001325865
inference algorithm 0.00132
feature func 0.001317344
text features 0.001296445
lexical features 0.0012961560000000001
novel algorithm 0.0012774269999999998
test set 0.001276265
beam search 0.001267457
learning task 0.001266829
deterministic algorithm 0.001256597
single learning 0.001255627
algorithm require 0.0012528519999999998
algorithm terminology 0.001247747
mira algorithm 0.001247747
markov model 0.001210475
learning strategy 0.001193261
learning framework 0.0011914109999999999
development data 0.001183078
tag set 0.0011815670000000001
other hypotheses 0.001168154
beam error 0.001165554
line learning 0.001156808
novel learning 0.001149015
search space 0.001136461
learning strategies 0.0011353959999999999
learning algo 0.001134722
supervised learning 0.001134283
other words 0.001134121
right search 0.0011316870000000001
learning phase 0.0011305669999999999
guided learning 0.001127059
aggressive learning 0.001122282
learning frame 0.0011212
context hypotheses 0.001120915
tron learning 0.0011206739999999999
sive learning 0.001119098
reinforcement learning 0.001119098
ment data 0.001117321
velopment data 0.00111658
training sequence 0.001099656
feature 0.00109004
search methods 0.001089927
bidirectional search 0.001085728
standard test 0.001081426
search strategy 0.001076259
label set 0.0010567760000000002
pos tag 0.00104471
possible pos 0.0010420770000000002
second set 0.001035743
tional search 0.0010295389999999999
algorithm 0.00102711
flexible search 0.001020939
deterministic search 0.0010111830000000001
top hypotheses 0.00101016
relative error 0.001004559
rectional search 0.0010022030000000001
standard hypothesis 9.82069E-4
pos tags 9.80273E-4
candidate hypotheses 9.77625E-4
standard action 9.73288E-4
new pos 9.71427E-4
same state 9.68702E-4
text hypotheses 9.64832E-4
multiple hypotheses 9.572890000000001E-4
error rate 9.50563E-4
standard sequence 9.47142E-4
action score 9.42462E-4
score action 9.42462E-4
features 9.37786E-4
linear function 9.28693E-4
empty set 9.26177E-4
large beam 9.21185E-4
same span 9.19334E-4
top hypothesis 9.190680000000001E-4
error rates 9.09464E-4
error reduction 9.087650000000001E-4
model 9.08089E-4
context information 9.05138E-4
error reduc 9.00758E-4
learning 8.98698E-4
