language model 0.00213742
model accuracy 0.001949101
model error 0.001871268
training data 0.0018186079999999998
different results 0.001777636
words precision 0.001746934
model output 0.0017304929999999998
model selection 0.0016604319999999999
model errors 0.00164737
word classes 0.001613731
character words 0.001605628
specific words 0.001580998
long words 0.001575347
words figure 0.001569908
test data 0.001563558
words increase 0.00155647
different number 0.001547515
model parameters 0.001545922
hidden words 0.001522863
matched words 0.00150138
related words 0.001498735
character word 0.001449528
data problem 0.0014093159999999999
different approaches 0.001398967
word segmentation 0.001381846
different lengths 0.001355588
matched word 0.00134528
different genre 0.0013426760000000001
word lengths 0.001341238
different assumptions 0.0013399969999999999
data entry 0.001326542
language models 0.001326009
word list 0.001325139
recognition accuracy 0.0013192770000000002
model 0.00129964
insufficient data 0.001278169
professional data 0.001276623
words 0.00124041
particular language 0.001231712
language processing 0.001209804
bigram language 0.001201951
chinese language 0.001164221
recognition approach 0.001163249
corresponding language 0.001160817
language properties 0.001160745
particular features 0.001156876
appropriate language 0.001139354
much language 0.001125375
individual features 0.00111128
asian language 0.001105008
specific features 0.001103532
recognition output 0.001100669
based language 0.0010945640000000001
complementary language 0.0010822100000000001
output accuracy 0.001080314
hybrid language 0.0010774790000000001
language rnoaei 0.0010774790000000001
single classifier 0.001071678
recognition improvement 0.001061665
detection accuracy 0.001061388
various features 0.001043233
error class 0.001036516
character recognition 0.001035034
tree classifier 0.00102218
high recognition 9.96717E-4
large set 9.91193E-4
error detection 9.83555E-4
classification performance 9.81721E-4
speech recognition 9.74003E-4
recognition techniques 9.5009E-4
coarse recognition 9.4992E-4
pattern recognition 9.408880000000001E-4
future work 9.32876E-4
single feature 9.29213E-4
recognition position 9.272900000000001E-4
error rate 9.26416E-4
descend method 9.2552E-4
following rule 9.24813E-4
handwriting recognition 9.159020000000001E-4
syntactic rules 9.15491E-4
recognition app 9.151420000000001E-4
handwritten recognition 9.1423E-4
recognition architecture 9.1095E-4
expensive recognition 9.1095E-4
text correction 9.04256E-4
low precision 9.01476E-4
manual error 8.988690000000001E-4
accuracy increases 8.93056E-4
measured accuracy 8.909780000000001E-4
detection performance 8.863079999999999E-4
same technique 8.838590000000001E-4
vector machine 8.799389999999999E-4
information 8.75979E-4
several classifiers 8.75189E-4
error correction 8.72864E-4
correct class 8.723800000000001E-4
character precision 8.71742E-4
entire text 8.68409E-4
precision overall 8.669799999999999E-4
post error 8.502480000000001E-4
