chinese transliteration 0.00503079
transliteration model 0.00374834
phonetic transliteration 0.0034162430000000002
transliteration models 0.003303355
semantic transliteration 0.003285836
same transliteration 0.003281193
chinese pinyin 0.0032574659999999997
chinese phoneme 0.0031738969999999997
chinese phonemes 0.0031263859999999997
chinese transliterations 0.003121764
chinese character 0.00309897
transliteration performance 0.003069421
chinese characters 0.003029961
chinese grapheme 0.0029695769999999997
chinese pronunciation 0.002968778
machine transliteration 0.002959394
chinese translit 0.002918582
correct transliteration 0.002906999
chinese homophone 0.002906046
hybrid transliteration 0.002898284
chinese graphemes 0.0028843109999999996
basic transliteration 0.002880197
english word 0.0028690300000000003
chinese transliter 0.002836421
chinese charac 0.002823952
transliteration mod 0.0027868240000000002
mantic transliteration 0.002777619
incorrect transliteration 0.002773215
english phoneme 0.002670407
english phonemes 0.002622896
english words 0.002581528
chinese 0.0025413
transliteration 0.00248949
corresponding english 0.002470574
english grapheme 0.002466087
ith english 0.002399838
multiple english 0.002392804
english graphemes 0.002380821
greeley english 0.002316284
english 0.00203781
phonetic translation 0.001724925
alignment results 0.001712932
corresponding model 0.001691614
entropy model 0.001672025
hybrid model 0.0016676439999999998
training data 0.0016534409999999999
class model 0.0016037599999999999
eration model 0.0015978519999999999
same training 0.001566792
model train 0.001564873
ation model 0.001563236
model gen 0.0015440419999999998
model inms 0.001536725
model parame 0.0015353189999999998
tropy model 0.0015353189999999998
exponential model 0.0015353189999999998
language phonemes 0.001508181
merged alignment 0.00140494
models transliterations 0.0013943290000000001
test data 0.001375263
models figure 0.00133238
phonetic translit 0.001304035
same test 0.001288614
same parameter 0.0012651000000000001
model 0.00125885
ing data 0.001250156
data size 0.00123555
hybrid models 0.001222659
basic models 0.001204572
word accuracy 0.00118725
pinyin table 0.0011777979999999999
development data 0.0011683050000000001
pinyin string 0.0011599219999999999
same example 0.001155861
eration models 0.001152867
tion models 0.0011389759999999999
glish word 0.0011337510000000001
andmj models 0.001132388
alignment 0.00112813
same con 0.0011200020000000002
ation models 0.001118251
phoneme research 0.001117822
word greeley 0.001109694
test set 0.001109177
themi models 0.001103058
thems models 0.001101652
themj models 0.001100298
conventional models 0.001098805
same question 0.001091965
models inms 0.00109174
models inmj 0.001089712
models independent 0.001089712
different sizes 0.001074352
other languages 0.0010732279999999999
same condition 0.001070185
personal names 0.001069283
nese pinyin 0.00106888
ent training 0.001064494
version method 0.001058513
person names 0.001054816
