language model 0.00375716
model features 0.0035066999999999997
translation model 0.003092615
error model 0.002839092
crf model 0.002625194
model size 0.002567288
transduction model 0.0025373590000000003
guage model 0.002534489
labeling model 0.002526308
word error 0.002523222
tive model 0.0025103720000000003
malization model 0.002510186
trained model 0.002508216
model variations 0.0025060110000000003
model normaliz 0.002504072
model versions 0.002504072
model mis 0.002504072
target word 0.0024977489999999996
word sequence 0.002344107
word level 0.002300516
model 0.00227895
language models 0.0021673639999999997
other language 0.002094364
training data 0.00193553
language analysis 0.001884813
single language 0.001841798
network language 0.0018206099999999998
srn language 0.0017752319999999999
dard language 0.0017593959999999999
unigram language 0.00173323
feature set 0.001682273
ing data 0.001668008
input features 0.0016020329999999999
text corpus 0.0015974729999999999
news text 0.00157344
text normalization 0.001563023
individual words 0.001546727
multiple words 0.001529091
srn features 0.0015247719999999998
translation models 0.0015028189999999999
twitter data 0.0014901839999999999
oov words 0.001482722
test data 0.001480447
all words 0.001479459
language 0.00147821
linguistic data 0.001468358
annotated data 0.001457623
novel words 0.001457383
features wer 0.001456757
syllable features 0.0014537999999999999
feature sets 0.0014474409999999998
random text 0.001402626
unlabeled data 0.00140243
neural text 0.001381078
development data 0.001367191
tweet text 0.0013225429999999998
parallel text 0.001313889
srn text 0.0013030619999999998
text representations 0.001279629
text message 0.00127691
text normaliza 0.001271726
additional training 0.001259332
text embeddings 0.001259079
novel text 0.001243423
scripts text 0.001238594
ral text 0.00123467
text writ 0.001232488
plausible text 0.001232488
plentiful text 0.001232488
balanced text 0.001232488
text segmen 0.001232488
features 0.00122775
words 0.00122
results table 0.001218948
feature 0.00119941
machine translation 0.001150431
target string 0.0011144599999999998
little training 0.001114145
statistical translation 0.001107975
evaluation results 0.001090853
source string 0.001087179
training examples 0.0010775
optimal results 0.001060003
labeled training 0.001057428
ing source 0.001044856
training srns 0.001030607
output string 0.001029854
normalization performance 0.001018217
following models 0.001008456
current time 9.793900000000001E-4
similar way 9.62933E-4
input string 9.54074E-4
normalization methods 9.50661E-4
such characteristics 9.47775E-4
guage models 9.446929999999999E-4
generic models 9.384829999999999E-4
previous time 9.37694E-4
error term 9.35624E-4
other tweet 9.32657E-4
large collection 9.216679999999999E-4
