first model 0.002353944
same model 0.00228302
hmm model 0.0022487419999999998
variable model 0.002210584
crf model 0.002083979
markov model 0.002065767
regression model 0.002056299
dplvm model 0.002050665
ing model 0.002038043
svm model 0.002015143
able model 0.001993904
model topology 0.001993065
model 0.001779
training data 0.001651737
other models 0.001415678
test data 0.0014103549999999999
raw data 0.001331926
data sparseness 0.001322218
severe data 0.001317382
data imbalance 0.001314326
english task 0.001309776
english abbreviation 0.001304585
chinese task 0.001302966
chinese abbreviation 0.001297775
word boundaries 0.00128124
labeling models 0.0012543480000000002
previous abbreviation 0.00125271
conditional probability 0.001247087
same feature 0.001237656
feature set 0.001226888
abbreviation characters 0.001217858
training corpus 0.00120249
training method 0.001185572
local information 0.001172368
variable models 0.001163867
generation task 0.001157802
possible feature 0.001153438
abbreviation generation 0.001152611
abbreviation process 0.001147995
other methods 0.0011418980000000001
feature generation 0.001139914
other baseline 0.001130984
generation performance 0.001127516
abbreviation length 0.001101548
simple feature 0.0010984789999999999
previous label 0.001095071
label set 0.0010819459999999999
abbreviation candidates 0.001071087
variable method 0.001061939
generation corpus 0.001053551
training set 0.001048469
conditional log 0.00104777
many label 0.001047297
training time 0.001042445
correct labeling 0.001030627
global information 0.001027348
additional information 0.001019494
various models 0.001019309
labeling problem 0.001015799
feature vector 0.001015563
other languages 0.001010501
recognition task 0.001001942
abbreviation recognition 9.96751E-4
abbreviation gener 9.9379E-4
same time 9.91248E-4
ferent task 9.91231E-4
detection features 9.91105E-4
optimization results 9.89577E-4
initial character 9.89001E-4
current character 9.878360000000002E-4
nese task 9.842380000000001E-4
feature templates 9.81752E-4
generation first 9.81222E-4
abbreviation definitions 9.78736E-4
feature design 9.778E-4
abbreviation generator 9.752179999999999E-4
glish abbreviation 9.72949E-4
sensitivity information 9.71485E-4
experimental results 9.70219E-4
abbreviation recognizers 9.67061E-4
abbreviation gen 9.618999999999999E-4
orthographic information 9.61601E-4
abbreviation recogni 9.61328E-4
abbreviation genera 9.61328E-4
abbreviation defini 9.60155E-4
abbreviation disam 9.60155E-4
abbreviation pga 9.60155E-4
appropriate features 9.53576E-4
feature tem 9.48477E-4
character duplication 9.44391E-4
chinese case 9.441E-4
generation figure 9.39336E-4
poor performance 9.38361E-4
other labelings 9.323000000000001E-4
encoding method 9.31826E-4
variable approach 9.27322E-4
independent features 9.25956E-4
sequence labeling 9.1379E-4
good results 9.126609999999999E-4
previous generation 9.126550000000001E-4
