different data 0.0025770719999999997
training data 0.002539292
data sets 0.002306824
twitter data 0.002302299
sms data 0.002297009
development data 0.002234206
data source 0.002186264
noisy data 0.002150751
particular data 0.002138803
data types 0.002119466
data sources 0.002084873
data styles 0.002082315
opment data 0.002080563
text normalization 0.0018567129999999999
channel model 0.0018518039999999999
generic model 0.0018067259999999998
based model 0.0017967629999999998
model formulation 0.001771953
statistical model 0.001757325
normalization performance 0.001713394
word tokens 0.0016672079999999999
word forms 0.001661824
word form 0.001647154
word error 0.001644708
normalization algorithm 0.00163091
word normalizations 0.001604934
model 0.00152686
word mappings 0.001522278
word annotations 0.001508178
word reordering 0.0015044589999999999
shorthand word 0.0014697859999999998
normalization work 0.001461024
normalization task 0.001438341
normalization problem 0.001425399
normalization graph 0.001399146
learning algorithm 0.0013862459999999998
graph models 0.0013857890000000001
feature set 0.001366603
twitter normalization 0.001363043
normalization approaches 0.001327233
feature functions 0.001303079
normalization process 0.001302085
normalization framework 0.001301429
similar models 0.001291351
normalization tasks 0.001287125
text message 0.001276004
normalization evaluation 0.001260807
message normalization 0.001252439
gold normalization 0.001243038
normalization gold 0.001243038
informal text 0.0012392710000000001
specific normalization 0.0012296619999999999
dictionary words 0.001216127
normalization techniques 0.001203788
raw text 0.0012031260000000001
certain normalization 0.001203114
correct text 0.001200592
typical normalization 0.00118026
reordering words 0.00116954
malized text 0.0011666
unnormalized text 0.001166263
clean text 0.001166153
mal text 0.001164321
actual normalization 0.001162674
normalization behavior 0.001155713
output sequence 0.001155453
standard forms 0.001153075
normalization actions 0.001139814
standard form 0.0011384049999999999
formed words 0.001136243
capitalized words 0.001136243
malize words 0.001136243
tion performance 0.001125042
input sequence 0.001106506
overall performance 0.001106155
parser performance 0.001105147
translation task 0.0010763600000000002
performance improvements 0.001073458
new domain 0.001070251
other case 0.001070106
gold standard 0.001062725
different domains 0.001060755
domain adaptation 0.001056543
sequence labeling 0.001043832
ization performance 0.001032329
absolute performance 0.001024473
previous work 0.0010240050000000001
normalizer performance 0.001023024
performance increases 0.001022765
performance mea 0.001022337
standard path 0.001000265
target language 0.001000036
standard annotations 9.99429E-4
partition function 9.95101E-4
jth feature 9.90979E-4
other approaches 9.90284E-4
algorithm path 9.783399999999999E-4
standard generation 9.7367E-4
different metrics 9.68741E-4
source language 9.67154E-4
