model model 0.00419076
language model 0.0032484000000000002
word sequence 0.003242134
word string 0.003163418
word annotation 0.0031614069999999998
word level 0.0031567749999999997
next word 0.003145182
word history 0.003117933
word predictor 0.0030788979999999996
word lat 0.003076088
word prefix 0.003075193
word dog 0.0030685799999999996
probabilistic model 0.002389943
entropy model 0.002376471
second model 0.002375272
our model 0.002309501
model actions 0.00230393
model halts 0.002296079
model 0.00209538
signs probability 0.00140432
training data 0.001312104
probability 0.00120441
single words 0.001195769
language 0.00115302
syntactic structure 0.001090064
irrelevant words 0.001086294
words 8.86582E-4
pos tag 8.82036E-4
possible parse 8.67561E-4
distance information 8.64556E-4
sentence boundary 8.51431E-4
yntactic structure 8.32436E-4
partial parse 8.20454E-4
degenerate approach 7.8394E-4
automatic speech 7.78294E-4
complete parse 7.60061E-4
speech recognizer 7.48969E-4
parsing process 7.24669E-4
perplexity results 6.94279E-4
parse note 6.88479E-4
joint sequence 6.84995E-4
distance dependencies 6.80366E-4
syntactic knowledge 6.72039E-4
long distance 6.5965E-4
first term 6.55369E-4
ditional probabilities 6.515080000000001E-4
feature 6.39436E-4
simple alternative 6.38627E-4
possible headword 6.37189E-4
meaningful information 6.34736E-4
structure 6.31163E-4
full power 6.29643E-4
algorithm 6.24343E-4
ing toolkit 6.19662E-4
level perplexity 6.06295E-4
bigram constraints 5.96688E-4
exposed head 5.965090000000001E-4
parser transition 5.92456E-4
unique sequence 5.91126E-4
constraint templates 5.81882E-4
approach 5.79998E-4
maximum entropy 5.76945E-4
sentence 5.66155E-4
last step 5.545249999999999E-4
exposed headword 5.53028E-4
headword annotation 5.512939999999999E-4
speech 5.477E-4
parser operation 5.45019E-4
training 5.42902E-4
present work 5.323050000000001E-4
work triggers 5.28427E-4
penn treebank 5.17062E-4
predictive power 5.15928E-4
models 5.1501E-4
exposed headwords 5.13832E-4
important ones 5.10474E-4
collins parser 5.06111E-4
topmost head 5.02751E-4
adjoin operations 4.886809999999999E-4
parse 4.88569E-4
boundary isnot 4.88523E-4
binary trees 4.8424899999999996E-4
function 4.77464E-4
main goal 4.7533700000000005E-4
mentary operations 4.73476E-4
iminary exper 4.73244E-4
exper iments 4.72032E-4
few provisions 4.68932E-4
binary parses 4.68319E-4
adjoin operation 4.66958E-4
headword assignments 4.6401099999999996E-4
equivalence classification 4.63347E-4
headword percolation 4.6179399999999996E-4
bilistic parametrization 4.5289E-4
dependency 4.39135E-4
distance 4.36658E-4
information 4.27898E-4
number 4.06388E-4
probabilities 3.87164E-4
process 3.79246E-4
