word corpus 0.003263169
language model 0.0029865
different word 0.002971703
new word 0.002858253
first word 0.002830415
last word 0.002716891
next word 0.002701286
current word 0.002678521
word probabili 0.002647143
fact word 0.002628739
subsequent word 0.002627784
word xul 0.002623725
word train 0.002602415
bayesian model 0.002323347
other words 0.00221716
different words 0.002161713
language models 0.002139048
guage model 0.0021004
tic model 0.002091496
perfect model 0.0020823639999999997
model selection 0.002063067
langauge model 0.002063067
test words 0.001991946
common words 0.001918321
model 0.00184141
unique words 0.001837193
rare words 0.001820546
subsequent words 0.001817794
words scales 0.001789413
drawing words 0.001788162
unigram language 0.001661581
bayesian language 0.001627027
words 0.00156522
natural language 0.0015279100000000002
probabilistic language 0.001517614
language processing 0.001509349
dirichlet language 0.0015067140000000001
bayesian models 0.001475895
good language 0.001464928
such models 0.001403144
novel language 0.001391861
language modelling 0.0013809170000000002
yor language 0.001373117
probabilistic models 0.001366482
training data 0.001286188
trigram models 0.001270396
complex models 0.001234361
dirichlet distribution 0.001217798
graphical models 0.0012171900000000002
posterior distribution 0.001184346
second probability 0.001164992
conditional distribution 0.001148958
language 0.00114509
test data 0.001144193
first probability 0.001130096
parent distribution 0.001116878
mean distribution 0.001100983
distribution gpi 0.001083123
different values 0.0010369160000000001
models 9.93958E-4
probability theory 9.83939E-4
training set 9.83282E-4
predictive probability 9.814119999999999E-4
other contexts 9.79035E-4
probability vectors 9.638159999999999E-4
data sets 9.617159999999999E-4
other variables 9.51957E-4
parameter values 9.46383E-4
probability estimates 9.409329999999999E-4
different discount 9.346300000000001E-4
new table 9.28936E-4
restaurant process 9.19627E-4
different problem 9.06504E-4
test time 9.05821E-4
dictive probability 8.988379999999999E-4
corpus 8.87959E-4
other hand 8.83814E-4
other benefits 8.74774E-4
form value 8.74101E-4
various values 8.67545E-4
dirichlet process 8.663970000000001E-4
distribution 8.56174E-4
bayesian approach 8.50324E-4
suffix tree 8.48625E-4
discount value 8.4672E-4
inference scheme 8.42211E-4
test set 8.41287E-4
large number 8.39578E-4
same values 8.38444E-4
hierarchical dirichlet 8.37845E-4
different ways 8.3622E-4
overall number 8.28657E-4
proper training 8.25247E-4
full training 8.21427E-4
different val 8.19425E-4
current number 8.07813E-4
new draw 8.05496E-4
first customer 8.029179999999999E-4
efficient inference 8.01778E-4
prior structure 7.92423E-4
