word features 0.00440095
test word 0.0029847520000000002
english word 0.002894472
word accuracy 0.002823088
rate word 0.002820501
word forms 0.002764449
complex word 0.002745837
word pronunciation 0.0027295
whole word 0.0026880050000000003
word accuracies 0.00267606
word overdo 0.00267606
word gorod 0.00267606
feature set 0.0023304709999999998
many features 0.002313684
additional features 0.002304858
contextual features 0.002271195
local features 0.0022563509999999998
suffix features 0.002231784
stress system 0.0022296720000000003
useful features 0.002214821
basic features 0.002214356
lexical stress 0.002204392
affix features 0.002191088
features gen 0.002170435
motivated features 0.002159369
fix features 0.002159369
possible stress 0.002087377
other words 0.002055772
primary stress 0.002033208
stress patterns 0.002016835
stress level 0.001991856
various feature 0.001972327
stress prediction 0.00197145
stress pattern 0.001955625
alternative stress 0.001955135
stress alternative 0.001955135
russian stress 0.001948922
features 0.00194222
stress symbol 0.00193045
secondary stress 0.00191084
complete feature 0.001900967
correct stress 0.001898596
feature combinations 0.001890051
ferent feature 0.001887817
sparse feature 0.001885093
feature combina 0.0018784729999999999
feature spaces 0.001877461
feature combi 0.0018763739999999999
incorrect stress 0.0018749840000000001
stress pat 0.001872013
stress assignment 0.001870435
invalid stress 0.001867887
valid stress 0.001867484
stress variants 0.001858016
ble stress 0.001851885
linear model 0.001844296
stress predictions 0.001839724
stress assignments 0.0018359820000000001
stress placement 0.001834821
stress possibilities 0.001833105
stress might 0.001833105
assigned stress 0.001833105
entropy model 0.0017654839999999999
feature 0.00165804
model com 0.001627803
novel words 0.001621749
stress 0.00161582
postaccented words 0.001613005
trained model 0.00157894
words 0.00137729
training data 0.0013495059999999999
model 0.00134776
morphological rules 0.00133439
data set 0.00131006
training approach 0.001190991
other work 0.001174179
test data 0.001163651
results table 0.0010877130000000001
other patterns 0.001079497
first syllable 0.001033322
efficient training 0.0010190849999999999
accuracy results 9.877900000000001E-4
machine translation 9.73422E-4
training examples 9.628799999999999E-4
sgd training 9.46894E-4
single test 9.46592E-4
data forms 9.43348E-4
ing data 9.36905E-4
training optimizer 9.29395E-4
positive set 9.16924E-4
same level 9.11383E-4
many speech 9.05589E-4
other hand 9.04019E-4
stressed syllable 8.88734E-4
same lemma 8.777030000000001E-4
our data 8.59086E-4
data our 8.59086E-4
umn results 8.43716E-4
automatic speech 8.38884E-4
phonetic class 8.19396E-4
