input word 0.004495265
output word 0.004162212
standard word 0.004125435
word order 0.004104072
word rule 0.004031642
word size 0.003987082
word constituent 0.003900058
word trigrams 0.0038964629999999997
word selection 0.003893745
head word 0.003886805
rightmost word 0.003882037
word choices 0.003875074
word choice 0.0038669
fixed word 0.0038509819999999998
leftmost word 0.003846348
word bigrams 0.003845473
word drops 0.003844947
word insertion 0.003842607
inal word 0.0038410429999999997
tent word 0.003840354
input words 0.0029535950000000003
extra words 0.002397565
single words 0.002383628
certain words 0.002350154
tion words 0.002319425
clude words 0.002306091
words 0.00207264
new model 0.001809657
grammar model 0.0017001490000000002
statistical model 0.001681083
input data 0.0016790210000000002
model parameters 0.0015948590000000001
input sentence 0.001578386
training data 0.001562324
linear model 0.001552913
guage model 0.0014936980000000001
feature vector 0.001481143
model grammati 0.001480753
level feature 0.001425749
language models 0.0013977590000000002
standard input 0.00139208
training algorithm 0.0013623189999999999
input set 0.001351123
time input 0.001345153
condition feature 0.001330154
search algorithm 0.001315735
feature constituent 0.001306518
different categories 0.001295021
global feature 0.001285886
input sentences 0.001271163
feature definitions 0.0012648120000000001
input length 0.001260841
model 0.00125766
feature template 0.001255614
perceptron training 0.001255486
feature tem 0.001251664
output sentence 0.001245333
learning algorithm 0.001244783
pos tag 0.001219661
input sequence 0.001214745
test search 0.001210482
different methods 0.001171494
generation features 0.001169429
first sentence 0.001162048
large search 0.00116158
improvement system 0.0011548349999999999
search problem 0.0011541049999999999
specific input 0.001152313
unique input 0.001144301
individual features 0.001126288
different timeout 0.0011204259999999999
full input 0.001113955
different sizes 0.001113294
training example 0.001113195
training process 0.001111682
current training 0.0011116379999999999
input check 0.001111073
input constraints 0.001110073
beam search 0.001109018
development data 0.0011072130000000001
noisy input 0.001106399
tary input 0.001105163
input multiset 0.001103992
input situation 0.001103992
input bag 0.001103992
data structures 0.0010955840000000001
experimental data 0.00108868
different numbers 0.001088417
corresponding sentence 0.001086915
different lengths 0.001080932
sentence length 0.001077317
search prob 0.001076002
reasonable training 0.001069675
search task 0.001065255
search process 0.001065098
first algorithm 0.001062678
training iterations 0.001042593
search space 0.001037956
complete sentence 0.001036744
training iteration 0.001035457
