language model 0.00159717
lexical data 0.001534823
word information 0.001528528
training data 0.00143748
first word 0.001330365
different training 0.001306996
current word 0.001280563
previous word 0.001273014
test data 0.0012457850000000001
same training 0.001207061
word pair 0.001123708
training corpus 0.001103848
lning data 0.001090087
word pak 0.001085117
model component 0.001063247
full parsing 0.001056031
other chunk 0.001053006
different context 0.0010475979999999998
speech information 0.001026274
different ways 0.001016504
different lexicons 9.93657E-4
lexical entry 9.84232E-4
other rule 9.79002E-4
contextual information 9.49634E-4
same evaluation 9.37347E-4
similar method 9.28839E-4
lexical entries 9.16955E-4
learning method 9.16402E-4
new tagger 9.148E-4
same parent 9.14032E-4
mutual information 9.004249999999999E-4
possible tag 8.96806E-4
current input 8.8456E-4
previous input 8.77011E-4
learning approach 8.76042E-4
error number 8.68842E-4
statistical models 8.638420000000001E-4
chunk tag 8.618079999999999E-4
first item 8.53716E-4
new lexicon 8.48404E-4
learning performance 8.46936E-4
new chunk 8.452430000000001E-4
model 8.31418E-4
corpus size 8.30735E-4
input string 8.26715E-4
grammar 8.20059E-4
corpus ize 8.13446E-4
training sections 8.13409E-4
chunk tagger 8.12969E-4
previous tags 8.066830000000001E-4
structural tag 7.97683E-4
optimal tag 7.94694E-4
current chunk 7.929650000000001E-4
tag sequence 7.88272E-4
single lexicon 7.81077E-4
words 7.78618E-4
language 7.65752E-4
corpus frequencies 7.62541E-4
previous part 7.60551E-4
chunk type 7.58368E-4
chain rules 7.565089999999999E-4
individual tag 7.48916E-4
learning technique 7.47966E-4
tag sequences 7.425159999999999E-4
tagging method 7.37989E-4
type overall 7.3779E-4
above chunk 7.25149E-4
new hidden 7.20696E-4
corresponding input 7.200780000000001E-4
large lexicon 7.17238E-4
accurate models 7.09908E-4
input tokens 7.09551E-4
text chunk 7.0931E-4
adjacent input 7.03738E-4
parsing 7.00977E-4
structural tags 7.00554E-4
syntactic structures 6.96398E-4
following order 6.939299999999999E-4
input oken 6.905E-4
baseline system 6.89503E-4
finite state 6.89287E-4
input okens 6.88787E-4
modelling approach 6.88549E-4
many researchers 6.8361E-4
several context 6.83074E-4
correct chunks 6.81168E-4
structural chunk 6.79287E-4
limited number 6.7922E-4
driven learning 6.76409E-4
viterbi tagger 6.72786E-4
second item 6.708129999999999E-4
information 6.69224E-4
lexicon memory 6.68391E-4
above experiments 6.65544E-4
main difference 6.647210000000001E-4
viterbi algorithm 6.63597E-4
state transducer 6.63552E-4
state transducers 6.63552E-4
type precision 6.63176E-4
english noun 6.612429999999999E-4
