language model 0.00486449
base model 0.0035061940000000002
generative model 0.003383891
model inference 0.003366982
graph model 0.0033658430000000003
model com 0.0033592960000000003
model parameters 0.003355368
mixture model 0.003327876
graphical model 0.0033205030000000003
guage model 0.003298644
refined model 0.003269382
model 0.00302497
language models 0.0027191890000000003
user language 0.002470896
document language 0.00234674
base language 0.002320744
query language 0.002188202
level language 0.002167746
mixture language 0.0021424260000000002
personalized language 0.002129158
unigram language 0.002124285
background language 0.002115699
tween language 0.002085107
other words 0.001852194
language 0.00183952
training data 0.001706884
similar words 0.0016266319999999998
training corpus 0.0016243170000000001
word frequency 0.001445309
stop words 0.001387601
base models 0.001360893
other variables 0.001309033
potential function 0.001219756
quality function 0.0011806030000000001
relationship function 0.001164584
unigram models 0.001164434
similarity function 0.001159756
friendship function 0.001154552
guage models 0.001153343
tial function 0.001152649
friend function 0.001151027
other smoothing 0.001150954
partition function 0.001148537
popularity function 0.001144502
same interpolation 0.00114294
same number 0.001142709
tential function 0.001141302
twitter data 0.001123644
words 0.00112076
testing data 0.001118892
experiment corpus 0.0011150280000000001
entire corpus 0.001105332
network information 0.001099155
probability distribution 0.001096265
target user 0.001043771
start user 0.001030341
same author 0.001024144
candidate variables 0.001019194
weight vector 0.001018833
feature functions 0.001012874
same category 0.001011153
random variables 0.001007344
other hand 9.811289999999999E-4
same person 9.71436E-4
interpolation method 9.62182E-4
probability mass 9.593539999999999E-4
smoothing method 9.558419999999999E-4
user lan 9.432970000000001E-4
local information 9.264519999999999E-4
user modeling 9.24501E-4
modeling user 9.24501E-4
personalized user 9.21014E-4
search space 9.18025E-4
unigram user 9.16141E-4
information retrieval 9.10073E-4
user relationship 9.01439E-4
function 8.94521E-4
few information 8.91593E-4
valid probability 8.88162E-4
models 8.79669E-4
user behavior 8.784940000000001E-4
unseen sentences 8.76012E-4
main difference 8.70635E-4
ful information 8.69493E-4
diverse information 8.69493E-4
main idea 8.64836E-4
relevant document 8.64109E-4
baseline methods 8.496840000000001E-4
main contributions 8.46655E-4
main chal 8.46655E-4
intrinsic evaluation 8.45385E-4
training 8.44005E-4
set perplexity 8.42509E-4
entire document 8.322399999999999E-4
attribute variables 8.31673E-4
such heuristics 8.28299E-4
evaluation metric 8.269190000000001E-4
dom variables 8.237610000000001E-4
extrinsic evaluation 8.115900000000001E-4
search algorithm 8.11462E-4
