language model 0.002632558
translation model 0.0024748590000000003
user model 0.0023390670000000002
model probabilities 0.002304039
model figure 0.002252711
linear model 0.002219809
entropy model 0.002187868
lation model 0.002114491
memd model 0.002101352
ibm model 0.002099894
combined model 0.002099894
model curve 0.002099894
model 0.00187469
word probabilities 0.001520979
word sequence 0.0014862439999999998
maximum word 0.001455608
next word 0.0014319539999999998
transtype data 0.001403541
word prefix 0.0013832189999999998
complete word 0.001381628
word predictions 0.0013698229999999998
word suffix 0.0013507859999999999
word sequences 0.001334226
empty word 0.0013259729999999998
training corpus 0.0013231430000000001
statistical models 0.001273341
search algorithm 0.0012563980000000002
able data 0.00119611
data col 0.001194509
other methods 0.0011564049999999999
lation models 0.001117638
first step 0.001105802
models supporto 0.001104207
channel models 0.001104207
test corpus 0.0010766299999999999
text prediction 0.001059825
different user 0.001042526
single words 0.0010315
training cor 0.001031213
language processing 0.001014274
other command 0.00100378
prediction system 0.001002785
statistical translation 9.95673E-4
other ways 9.905369999999999E-4
trigram language 9.84273E-4
ing times 9.79843E-4
empirical probability 9.66253E-4
large corpus 9.49869E-4
benefit function 9.48701E-4
machine translation 9.4126E-4
maximum time 9.321760000000001E-4
time maximum 9.321760000000001E-4
new approach 9.151459999999999E-4
different predictor 9.1278E-4
current prediction 9.089359999999999E-4
basic prediction 9.04302E-4
french translation 8.96867E-4
first implementation 8.90343E-4
random variable 8.88755E-4
models 8.77837E-4
average time 8.72377E-4
target words 8.682220000000001E-4
standard user 8.63841E-4
first approximations 8.53646E-4
acceptance probability 8.523819999999999E-4
adjacent words 8.477070000000001E-4
next step 8.465759999999999E-4
first point 8.43764E-4
hansard corpus 8.41758E-4
typing time 8.38126E-4
objective function 8.380010000000001E-4
corpus perplexity 8.362160000000001E-4
empirical search 8.305070000000001E-4
ulated results 8.28567E-4
real time 8.27805E-4
translation technology 8.26456E-4
different domain 8.20383E-4
search problem 8.16124E-4
different trans 8.14252E-4
prediction task 8.13622E-4
prediction prototype 8.073519999999999E-4
prototype prediction 8.073519999999999E-4
user benefit 8.07212E-4
total time 8.065769999999999E-4
different compo 8.057990000000001E-4
next user 8.047009999999999E-4
predictor performance 8.005600000000001E-4
prediction proto 7.94674E-4
didate prediction 7.94674E-4
above example 7.94281E-4
sequence length 7.90935E-4
corresponding user 7.86914E-4
algorithm 7.74418E-4
average user 7.68556E-4
empirical methods 7.64349E-4
probabilistic acceptance 7.6412E-4
true value 7.6354E-4
main problems 7.62998E-4
user interface 7.61957E-4
editing method 7.60377E-4
