language model 0.00402519
word models 0.003348311
training data 0.0032932
unsupervised word 0.0030815060000000003
model figure 0.0030554939999999997
baseline model 0.0030248569999999997
word generation 0.003012798
oov word 0.002981632
new word 0.002972024
affix model 0.0029622529999999998
expansion model 0.002953204
word list 0.0029141870000000004
word letters 0.0028993390000000003
word func 0.002853243
word rate 0.0028499880000000003
wfst model 0.0028485059999999998
sion model 0.002841639
word recognition 0.0028336420000000004
language training 0.0028299700000000002
wrong word 0.0028257290000000003
inv word 0.002825465
model expanded 0.00281973
word combinations 0.0028162390000000003
word producttions 0.0028034
word combi 0.0028034
model 0.002535
evaluation data 0.002329005
language models 0.002313771
annotated data 0.002301821
project data 0.002286063
data type 0.002284696
development data 0.00226368
data development 0.00226368
data the 0.002262184
data sets 0.002257019
pack data 0.002239866
optical data 0.002236363
unannotated data 0.002236237
required data 0.002235924
data points 0.002233811
target language 0.002080952
morphological models 0.002054021
morphological information 0.0019992639999999997
training corpus 0.001893359
limited language 0.001882614
full language 0.00185585
human language 0.0018202749999999999
trigraph language 0.001815201
bigram language 0.00178197
language pack 0.0017766359999999998
language vocab 0.001771037
language technolo 0.0017679319999999998
translation models 0.001740114
different models 0.001718569
full training 0.0017054400000000001
annotated training 0.001688181
morphological segmentation 0.001682811
unlabeled training 0.001648105
example training 0.001637365
pack training 0.0016262260000000001
training transcripts 0.0016205450000000001
morphological modeling 0.001610809
other words 0.001596304
morphological analyzer 0.001583613
morphological mapping 0.0015778279999999999
morphological annotations 0.001555314
morphological complexity 0.001539073
morphological properties 0.001517799
morphological analyz 0.001509585
language 0.00149019
backoff models 0.001447981
oov words 0.0013920920000000002
new words 0.001382484
source words 0.00138063
morphology models 0.0013619510000000001
reranking models 0.001344294
training 0.00133978
unseen words 0.0013335880000000001
english words 0.001320194
other languages 0.0013199129999999998
morphology information 0.001307194
english translation 0.001301537
different morphemes 0.001288552
segmented words 0.0012841039999999999
relevant words 0.001243026
different ways 0.001231175
possible trigraphs 0.0012113129999999999
different sizes 0.001206401
different techniques 0.001194001
first stem 0.001192739
first suffix 0.001190013
character set 0.001182554
different thresholds 0.001174857
possible segmentations 0.00116116
first step 0.001152433
large number 0.00113728
possible metric 0.001125959
possible pronunciations 0.001125092
possible reason 0.001122773
possible transliterations 0.001122773
