language model 0.00218499
transition model 0.0017462149999999997
model automaton 0.001726196
new language 0.0016046580000000001
backoff model 0.00160152
input string 0.001570957
first state 0.00156635
exact model 0.0015400779999999998
language models 0.00151365
probabilistic model 0.0014958249999999999
weighted automaton 0.001491933
statistical language 0.0014836340000000002
new state 0.0014713299999999999
gram model 0.0014657779999999998
weighted transducer 0.0014626420000000001
interpolation model 0.001458512
trigram model 0.001458361
language processing 0.001453807
input speech 0.0014466230000000002
transition set 0.001439503
natural language 0.001432525
large set 0.0014066880000000001
bigram model 0.0014005049999999998
general grammar 0.001393142
backoff state 0.001381562
gram language 0.001379148
model building 0.0013791279999999999
grammar processing 0.001375825
classical language 0.001355988
language modeling 0.001345431
weighted grammars 0.00133919
grammar software 0.001337765
finite set 0.00132721
previous state 0.001323791
language process 0.001311404
language mod 0.0013094860000000001
large input 0.0013012800000000001
next state 0.00129365
initial state 0.00127689
speech recognition 0.0012562760000000002
final state 0.001246529
simple algorithm 0.001236704
input word 0.001235665
such models 0.0012328999999999999
finite input 0.0012218020000000001
weighted automata 0.001217617
efficient algorithm 0.001217361
following set 0.001208369
weighted difference 0.001198461
start state 0.001196485
weighted transducers 0.0011947020000000002
machine translation 0.001188356
other systems 0.001177375
probability semiring 0.001171848
destination state 0.001165356
counting algorithm 0.001165105
nation state 0.001156306
cost path 0.001153358
weighted transduction 0.0011469380000000001
weighted automa 0.001143458
removal algorithm 0.001138365
model 0.00113581
speech processing 0.00112756
conditional probability 0.001125025
such classes 0.0011194859999999998
string labeling 0.001100189
input corpus 0.00109503
finite automaton 0.001088498
infinite set 0.001087226
failure transition 0.001080114
recognition system 0.001077837
backoff states 0.001074468
batch set 0.001066738
weight function 0.001064175
transducer case 0.0010528550000000001
language 0.00104918
probability distribution 0.001048607
possible states 0.001048429
exact representation 0.001046694
recognition systems 0.001044331
other text 0.001043535
large models 0.00104206
likelihood probability 0.001041869
probability distributions 0.001036427
new method 0.001030918
second path 0.001028117
natural representation 0.001025771
start symbol 0.001014231
joint probability 0.001013643
input label 0.001013113
transition weights 0.001003543
end symbol 9.99437E-4
input alphabet 9.99429E-4
speech recognizer 9.92732E-4
particular input 9.91549E-4
offline representation 9.901699999999999E-4
speech utterances 9.88423E-4
speech process 9.85157E-4
general approach 9.8475E-4
alternative transition 9.84323E-4
