word alignment 0.003962810000000001
amr word 0.0030330170000000003
english word 0.0029725800000000003
training model 0.002827407
training alignment 0.002801177
ibm word 0.002763862
alignment models 0.002670744
word align 0.002565691
ibm model 0.002557662
alignment problem 0.002553272
test model 0.0025506599999999997
alignment method 0.002488011
word sense 0.002460466
stop word 0.002428978
word lists 0.0024226350000000002
word thanks 0.002414912
gold alignment 0.0023875980000000003
generative model 0.002330517
alignment links 0.002323518
alignment con 0.002291347
model precision 0.002274218
dev model 0.0022675539999999998
string alignment 0.002264017
alignment accuracy 0.00223514
translation table 0.0021948840000000002
alignment parame 0.002185793
machine translation 0.0021196690000000002
such alignments 0.001986616
training algorithm 0.001909656
model 0.00189142
alignment 0.00186519
gold alignments 0.0018515180000000001
amr concept 0.001741251
english words 0.001735477
automatic alignments 0.001681804
matic alignments 0.001648055
english sentence 0.001644353
amr graph 0.0016368580000000001
translation 0.00162523
training method 0.001558808
first search 0.001551752
training pairs 0.0015473470000000001
amr tokens 0.001520469
ibm models 0.001471796
english tokens 0.001460032
example english 0.001460006
english sentences 0.0014348120000000002
amr concepts 0.001418136
amr role 0.001387784
role amr 0.001387784
sentence pairs 0.0013807530000000002
viterbi algorithm 0.001372998
objective function 0.001369692
amr variables 0.0013606030000000002
test set 0.001357741
first set 0.001356994
data description 0.001352045
probability links 0.001340831
amr structures 0.001331868
strings data 0.001331125
alignments 0.00132911
whole data 0.00132887
amr graphs 0.001328521
full amr 0.001311368
amr node 0.001303727
amr structure 0.0012931190000000001
training phase 0.001287363
amr ones 0.001276738
multiple amr 0.001273843
available training 0.001266535
only english 0.0012581930000000001
ment models 0.001256565
linearized amr 0.001253261
unordered amr 0.001253261
amr design 0.001253261
different models 0.0012503710000000001
first step 0.001248029
high probability 0.001244326
negative words 0.00124344
sentence pair 0.001231913
parallel corpus 0.00122862
english ones 0.001216301
fertility table 0.001214275
pair corpus 0.001210526
english sides 0.001194231
stop words 0.001191875
insert words 0.00118492
smt work 0.001175334
ment set 0.001149512
generative graph 0.001140558
such links 0.001115834
test role 0.001111627
symmetric problem 0.001110552
future work 0.0010838129999999999
other information 0.001077913
test sets 0.001072724
token pairs 0.00107062
graph structure 0.001059183
symmetric method 0.0010452909999999998
role tokens 0.001037459
